The Evolution of Intumescent Char in Flame-Retardant Coatings Based on Amino Resin

Round 1

Reviewer 1 Report

Dear Authors,

the manuscript entitled "The Evolution of Intumescent Char of Flame-retardant Coatings Based on Amino Resin” by Wei Song, Muting Wu, Yanrong He, Yuzhang Wu and Wei Qu investigates the use of IFR made of melamine modified urea-formaldehyde resin, ammonium polyphosphate and pentaerythritol. The flame-retardant properties of the coating were measured by the cone calorimeter and the evolution of the intumescent char was studied.

The article has an interesting and well developed focus. In my opinion only small variations are necessary for the publication.

It seems that other approaches to depositing a protective layer on different substrates have already been applied in the literature, probably it is necessary to expand the references in the introduction e.g. “Bulk vs. surface flame retardancy of fully bio-based polyamide 10,10. Rsc Advances. 2015;5(49):39424-32.”

How many tests have been carried out at the cone calorimeter? Is the reported data an average? Is there a standard deviation of the reported data? As is well known, the calorimetric cone tests are affected by a certain uncertainty in the measurement and this must be evaluated. If the results refer to a single sample it must be repeated since it is the basic result on which the whole article is based.

FTIR analyzes for proper comparison should be normalized, has this been done?

In figure 8 I recommend putting marks for the peaks that are described in the text to make them more easily identifiable.

Author Response

1. It seems that other approaches to depositing a protective layer on different substrates have already been applied in the literature, probably it is necessary to expand the references in the introduction e.g. “Bulk vs. surface flame retardancy of fully bio-based polyamide 10,10. Rsc Advances. 2015;5(49):39424-32.”

Answer: Thanks a lot for the reviewer’s good reminder. This paper mentioned is very important to our research and was inserted as No. 7 reference.

2. How many tests have been carried out at the cone calorimeter? Is the reported data an average? Is there a standard deviation of the reported data? As is well known, the calorimetric cone tests are affected by a certain uncertainty in the measurement and this must be evaluated. If the results refer to a single sample it must be repeated since it is the basic result on which the whole article is based.

Answer: We greatly appreciated the reviewer’s question. Three replications were performed for each MDF types. The results of the three replications are similar. The influence of a certain uncertainty in the measurement is evaluated. The HRR and THR curves of MDF are discussed in this paper, so the standard deviation of the reported data is not provided in this article. Thus, the corresponding sentence in Materials and Methods have been changed to “The flame-retardant performance of MDF coated with A3P1 resin is tested by the cone calorimeter (CONE, FTT) according to ISO 5660-1 [23] under an external heat flux of 50 kW/m². Three replications were performed for each MDF types.”.

3. FTIR analyzes for proper comparison should be normalized, has this been done? In figure 8 I recommend putting marks for the peaks that are described in the text to make them more easily identifiable.

Answer: We greatly appreciated the reviewer’s excellent reminder and recommendation. The data of FTIR has been normalized before analysis. As the reviewer said, the marks for the peaks is necessary to identify them easily in Figure 8. Thus, Figure 8 have been replaced by a new figure with marks.

Reviewer 2 Report

The authors present a detailed analysis of the intumescent char obtained from a coating consisting of ammonium polyphosphate, pentaerythritol and melamine (APP-PER-MEL) in a melamine modified urea-formaldehyde resin (MUF) on steel. The topic is appropriate for the audience of Coatings. However, the paper lacks in high originality and significance from a flame retardancy point of view. The reader can follow the authors very well and the methods were suitable for this research. The authors should at least clarify the following remarks, before the paper can be published:

a There are some deficits in the English language and some careless mistakes (e.g. l.87: stell l.93 Steel, l.151 there,….). Additionally, there are many space characters missing between values and units (e.g. l. 77, ll. 94-95...).

b What is the thickness of the coatings and is it appropriate? Is there any reference for typical coating thicknesses used in practice (for steel and MDF)?

c Flame retardant properties: Values for TTI and time to flameout should be shown in a table, pHRR and THR(600 s) could be included as well. Are there any observations for char height/volume or the evolution of char vs. time for the different coating weights? It should be included, that the heat flux of the cone heater is higher at lower distances to the heater influencing the char formation and decomposition.

d Choice of substrate:

1. The substrate choices for the different tests should be addressed anywhere.
2. The flame retardant properties for this coating were shown on MDF but the formation of char was investigated on steel. Are there any references or observations that the substrate has no influence on the formation of char and the results can be transferred? Is there any cone calorimetry data of the coating on steel for comparison or char volume on MDF vice versa?

e Thermal analysis:

1. 158 The anorganic network should be specified, if the reference gives information about it.
2. For PER: Is it sure, that the TGA curve shows the decomposition and that the molecules are not transferred to the carrier gas undecomposed?

f Volumes of intumescent chars:

1. The choice of 800 °C should be explained in a sentence.
2. Maybe, a short comparison to observations of cone calorimetry measurements could be made.
3. How does the thermal conductivity correlate to the size and location of the pores? A brief description would be helpful to the reader.

g FTIR: Fig 8 – IR spectra usually go from higher to lower wavenumbers.

Author Response

1. There are some deficits in the English language and some careless mistakes (e.g. l.87: stell l.93 Steel, l.151 there,….). Additionally, there are many space characters missing between values and units (e.g. l. 77, ll. 94-95...).

Answer: We greatly appreciated the reviewer’s recommendation and question. We apologize for these careless mistakes. These mistakes including these mentioned by the reviewer have been modified in the revision.

2. What is the thickness of the coatings and is it appropriate? Is there any reference for typical coating thicknesses used in practice (for steel and MDF)?

Answer: Thanks a lot for the reviewer’s question. The thickness of the coating layer is very important for the application. Commonly, the thickness of the fire-retardant coating is higher than that of the ordinary coating in order to achieve a certain level of flame retardancy. The thickness of the coatings increases with the coating weight rising. When the coating weight of A3P1 is 250 g/m², the thickness of the dry layer is about 0.3 mm which is appropriate for most wood-based products. According to the thickness, the fire-retardant coating for steel is divided into 3 types as follows: < 3mm, ultra-thin; 3-7 mm, thin; >7mm, thick.

3. Flame retardant properties: Values for TTI and time to flameout should be shown in a table, pHRR and THR (600 s) could be included as well. Are there any observations for char height/volume or the evolution of char vs. time for the different coating weights? It should be included, that the heat flux of the cone heater is higher at lower distances to the heater influencing the char formation and decomposition.

Answer: We greatly appreciated the reviewer’s recommendation and question. A table of the cone calorimeter date included the TTI, pHRR, THR, and the like is very useful to compare the fire retardant properties of samples. When the difference is not very significant, the values of the parameters is more important. From Figure 1, there is a clear difference between the HRR curves of MDF coated with various coating weight. For example, the pHRR of MDF is more than 250 kW/m² while the peak is disappeared with coating weight more than 250 g/m². The maximum of char value increases when the coating weight is higher. However, the evolution of char is similar. The influence of the heat flux and the distance is mainly sourced from the heat temperature. When the temperature is higher, the speed of the evolution will increase. The coating weight chosen in this article is due to the actual application for Class II. The temperature is chosen by the data of TGA and the heat flux of cone calorimeter.

4. Choice of substrate: 1) The substrate choices for the different tests should be addressed anywhere. 2) The flame retardant properties for this coating were shown on MDF but the formation of char was investigated on steel. Are there any references or observations that the substrate has no influence on the formation of char and the results can be transferred? Is there any cone calorimetry data of the coating on steel for comparison or char volume on MDF vice versa?

Answer: We greatly appreciated the reviewer’s recommendation and question. 1) The substrate of samples for the cone calorimeter is MDF while the substrate of samples to obtain the char for subsequent characterization is steel. To address the substrate, the corresponding sentence in Materials and Methods have been changed to “Fourier transform infrared spectra (FTIR) of A3P1 resin and char prepared from steel samples are recorded by using Nicolet Is50 from 400 to 4000 cm^-1. The rate of the differentiate is 4 cm^-1 and the number of scanning is 8.”.  2) After the formation of char, there is a temperature gradient from the top of char to the interface of the coating and MDF. As we all know, the MDF or other wood-based materials are stable below ca. 200 ^oC. So there will no volatiles from MDF which can change the char formation and shrinkage. When the char cracks, the temperature of MDF increases. The volatiles from MDF can be flamed, which leads to the higher temperature. The destruction of char is accelerated. However, the steel substrate has on influence in the char evolution. It is a little difficult to obtain the char samples at various time under the cone calorimeter. Thus, the steel sample is alternative method to study the char formation and shrinkage of the fire-retardant coatings.

5. Thermal analysis: 1) 158 The anorganic network should be specified, if the reference gives information about it. 2) For PER: Is it sure, that the TGA curve shows the decomposition and that the molecules are not transferred to the carrier gas undecomposed?

Answer: We greatly appreciated the reviewer’s recommendation and question. 1) The third peak of weight loss is ascribed to the char oxidation. After this process, only residue char and inorganic substances were retained. Main inorganic substances are pyrophosphoric acid, according to the No.19 reference. This statement might not be accurate. Thus, the corresponding sentence have been changed to “The third peak is about the inorganic substance formation (pyrophosphoric acid, etc.) with 13% weight loss.”. 2) The TGA curve of PER is correct, because we repeat it three times which shows the same results. This result is consistent with previous studies.

6. Volumes of intumescent chars: 1) The choice of 800 °C should be explained in a sentence. 2) Maybe, a short comparison to observations of cone calorimetry measurements could be made. 3) How does the thermal conductivity correlate to the size and location of the pores? A brief description would be helpful to the reader.

Answer: We greatly appreciated the reviewer’s recommendation and question. 1) The explanation of the choice of 800 ^oC is necessary. The choice of 800 °C is according to the TGA results. A3P1 can be decomposed completely at 800 ^oC. And the temperature is close to the heat temperature of the cone calorimeter at 50 kW/m². Thus, the corresponding sentence have been changed to “The whole evolution of char both char formation and shrinkage was monitored at 800 ^oC in the muffle furnace with the coating weight 333 g/m². According to the TGA results, A3P1 could decompose completely at 800 ^oC.”. 2) It is difficult to get the detail information of char under the cone calorimeter. The char formation and shrinkage are similar with the char in muffle furnace. 3)The relationship between the thermal conductivity and the microstructure of char have not been established. The thermal conductivity of char is macroscopic results generally. It is our next research focus.

7. FTIR: Fig 8 – IR spectra usually go from higher to lower wavenumbers.

Answer: We greatly appreciated the reviewer’s recommendation. The main differences of IR spectra in this article are in the lower wavenumbers, so the spectra go from lower to higher. And the marks of peaks are added identify them easily.

Reviewer 3 Report

In this manuscript, Song et al. introduce a comprehensive and detailed analysis of the evolution of the structural, morphological and chemical properties of intumescent char in IFR coatings based on amino-resins. The novelty of the work is given by the three stages authors identified in the evolution process of the intumescent char, which could be profitably used for improving the flame-retardant properties of the coatings.

The subject is very-well introduced by the authors in the introductory section, providing the reader with all the info necessary to understand the aim and the context of the work. Previous literature is adequately cited.

The research design is appropriate, as well as the description of the methodology used. Conclusions are well supported by the results obtained with several different measurement techniques (cone calorimetry, SEM and FTIR).

I think that the manuscript quality, in terms of novelty and scientific soundness, is good enough to be considered for publication in Coatings journal.

However, in my opinion, the paper should be revised before publication, mainly because authors should add some considerations on how their results can be used for a possible upgrade of IFR coatings. I mean, all the obtained results (CONE, SEM, and FTIR) confirm the 3-stage evolution of the intumescent char, but the description of a strategy on how in practice this evolution can be used for improving flame-retardant properties of the resulting coating is missing. For instance, authors write (line 177)  “We can infer that the decrease of the shrinkage in the second stage can may be a new method to improve the flame-retardant properties of coatings”. But how? By increasing the thickness of the coating, for example? Have the authors identified an optimal thickness? Or by modifying the composition of the used APP + PER + MUF mixture? I invite the authors to elaborate a little bit more on this.

Herewith a detailed line-to-line list of other minor issues to be addressed:

• Line 15. Replace “Fourier-transform infrared spectra” with “Fourier-Transform Infrared Spectroscopy”, which is the correct extended name of FTIR technique.
• Line 25. Define melamine (MEL) after ammonium polyphosphate (APP), and pentaerythritol (PER) for consistency with the order of the acronyms (APP-PER-MEL).
• Line 45. Authors write “The flame-retardant performance of the mixture of APP and MUF was studied in our previous work”. Authors are invited to resume in a few words the results of this characterization performed in Ref. [12], for completeness.
• Line 68. Define “n” and “M_n”.
• Line 98. Remove the phrase “SEM is used at high vacuum conditions”, because it really sounds trivial. SEM works indeed only at UHV conditions, otherwise the electron beam cannot even be formed.
• Line 104. Replace “dimeter” with “diameter”.
• Line 106. I would remove the phrase “LED light adjusted to level 5 to maintain the proper and stable light condition”. I mean that “level 5” is not a standard brightness value, which is different from LED to LED. So the info on the LED light brightness should be removed or, better, the maximum brightness level (5 itself?) of the LED used should be explicitly written.
• Line 133. Authors write that “According to Building Standard Law of Japan, the samples with coating weight more than 250 g/m² satisfy Class II, semi - noncombustible material”. So do they infer that 250 g/m² is an optimal coating weight? Or do they think that 400 g/m² is better? Does 250 g/m² represent a good trade-off value between performance and weight? Authors are invited to discuss a little about this.
• Figure 2b. DTG peaks of A3P1 are not well-highlighted in the figure and almost indistinguishable from the APP-PER-MUF ones. I would show them in a separate figure, or an inset, for sake’s of clarity.
• Line 158. The range of the third peak is missing. Please add it in the revised version of the paper.

Author Response

1. Line 15. Replace “Fourier-transform infrared spectra” with “Fourier-Transform Infrared Spectroscopy”, which is the correct extended name of FTIR technique.

Answer: Thanks a lot for the reviewer’s good reminder. The corresponding word have been replaced in the revision.

2. Line 25. Define melamine (MEL) after ammonium polyphosphate (APP), and pentaerythritol (PER) for consistency with the order of the acronyms (APP-PER-MEL).

Answer: Thanks a lot for the reviewer’s good reminder. The corresponding word have been changed in the revision.

3. Line 45. Authors write “The flame-retardant performance of the mixture of APP and MUF was studied in our previous work”. Authors are invited to resume in a few words the results of this characterization performed in Ref. [12], for completeness.

Answer: We greatly appreciated the reviewer’s recommendation. It is necessary to add some words about the results of our previous work. Thus, the corresponding sentence have been changed into “The flame-retardant performance of the mixture of APP and MUF was studied in our previous work [13]. The flame-retardant performance of samples is related with the amount of APP closely. To improve its flame-retardant performance and lower the cost, the coating containing APP, PER, and MUF should be prepared and its mechanism needs to be fully investigated.”.

4. Line 68. Define “n” and “M_n”.

Answer: Thanks a lot for the reviewer’s good reminder. “n” is the polymerization degree. “Mn” is relative molecular weight. The corresponding word have been changed into “Ammonium polyphosphate (APP, n (degree of polymerization) > 1000) was purchased from Sichuan Changfeng Chemical Co., Ltd.”.

5. Line 98. Remove the phrase “SEM is used at high vacuum conditions”, because it really sounds trivial. SEM works indeed only at UHV conditions, otherwise the electron beam cannot even be formed.

Answer: Thanks a lot for the reviewer’s good reminder. The corresponding word have been deleted in the revision.

6. Line 104. Replace “dimeter” with “diameter”.

Answer: Thanks a lot for the reviewer’s good reminder. The corresponding word have been changed in the revision.

7. Line 106. I would remove the phrase “LED light adjusted to level 5 to maintain the proper and stable light condition”. I mean that “level 5” is not a standard brightness value, which is different from LED to LED. So the info on the LED light brightness should be removed or, better, the maximum brightness level (5 itself?) of the LED used should be explicitly written.

Answer: We greatly appreciated the reviewer’s recommendation. The corresponding word have been deleted in the revision.

8. Line 133. Authors write that “According to Building Standard Law of Japan, the samples with coating weight more than 250 g/m² satisfy Class II, semi - noncombustible material”. So do they infer that 250 g/m² is an optimal coating weight? Or do they think that 400 g/m²is better? Does 250 g/m²represent a good trade-off value between performance and weight? Authors are invited to discuss a little about this.

Answer: We greatly appreciated the reviewer’s recommendation and question. Class II, semi-noncombustible material is the common requirement of wood-based product. It indicates that 250 g/m² represent a good trade-off value between performance and weight.

9. Figure 2b. DTG peaks of A3P1 are not well-highlighted in the figure and almost indistinguishable from the APP-PER-MUF ones. I would show them in a separate figure, or an inset, for sake’s of clarity. Line 158. The range of the third peak is missing. Please add it in the revised version of the paper.

Answer: We greatly appreciated the reviewer’s recommendation. The range of the third peak is added in the revision.