Recycling Polyethylene Terephthalate (PET) Bottle Waste into Sustainable Building Materials: A Systematic Review of Technical Performance, Environmental Safety and Circular-Economy

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The review concerning "Recycling Polyethylene terephthalate (PET) Waste in making Building Materials" was reviewed. Although the topic is interesting and advancing in specific research fields, there are still several contents need to be revised before accepted to publishing.

1. Contents in Introduction section should not to be divided into different parts, which is not meaningful for readers to understanding the research status for PET recycing in making building materials.
2. Format of tables and figures should be modified to ensure its compliance with academic standards and aesthetic appeal.
3. The references lack support from industry research reports or news. Are there any practical application cases for this technology? It is recommended to be clarified in Introduction section.
4. Section 4: The content in this section is divided into too many sub-chapters, and most of the sub-chapters lack in-depth analysis and discussion. Also, many data lack literature support.

Author Response

Thank you the editor and all reviewers for the careful reading of the manuscript and for the constructive comments, which have substantially improved the clarity, reproducibility and evidential basis of this review. We appreciate in particular the identification of inconsistencies between the abstract and the methodology, the lack of an appraisal of study quality, and the problems in the reference list; these have all been addressed directly. A point-by-point response is given below. All changes are shown in the revised manuscript.

Comment 1.1

Contents in Introduction section should not to be divided into different parts, which is not meaningful for readers to understanding the research status for PET recycing in making building materials.

Response: We thank the reviewer for this observation and agree. The three part-headings in the Introduction have been deleted and the material has been re-sequenced as one continuous argument that moves from PET production and waste generation, through the construction-sector context, to the research gap and the objectives of this review.

Revised text: the headings "1.1. Significance of PET in the Construction Industry", "1.2. Research Gaps" and "1.3. Scope and Objectives of the Current Review" have been removed from Section 1, and the Introduction now runs as a single narrative.

Comment 1.2

Format of tables and figures should be modified to ensure its compliance with academic standards and aesthetic appeal.

Response: We thank the reviewer. All figure and table captions have been rewritten to follow a consistent structure: what is shown, how the values were derived, the units, the data sources and, where relevant, the limitations of the comparison. Table 2 now reports an explicit dosage basis for every material class, and the note beneath it states that dosages expressed on different bases are not comparable. The captions of Figures 3, 4 and 6 now state the derivation of the plotted values and the normalisation applied, and the per-source values are provided in the Supplementary Materials.

Revised caption (Figure 3): "Bars show the midpoint of the ranges reported by the individual sources listed in Table 1, where the midpoint is the arithmetic mean of the reported lower and upper bound. Unit conversions and the normalisation procedure are described in Supplementary Table S3, together with the per-source values. The panels are indicative only: cost and energy values are not normalised to a common functional unit, so the panels should not be read as a like-for-like ranking."

Revised caption (Table 2): "Reported mechanical and thermal properties of PET-modified construction materials. The dosage basis is stated explicitly for each material class, because the values are expressed on different bases and are not interchangeable."

Comment 1.3

The references lack support from industry research reports or news. Are there any practical application cases for this technology? It is recommended to be clarified in Introduction section.

Response: We thank the reviewer for this suggestion, which has been adopted. A new paragraph at the end of the Introduction describes practical and field-scale adoption and cites industry and policy evidence alongside the peer-reviewed literature. Three sources have been added: a field-performance evaluation of recycled plastic in asphalt mixes, a detailed account of the volumetric and mechanical behaviour of waste-plastic-modified asphalt mixtures, and the OECD Global Plastics Outlook. A documented case of plastic-bonded sand blocks is cited from a non-PET polymer and is explicitly labelled as background evidence, in line with the distinction requested in review comment 2.19.

Added text (Section 1): "Practical adoption is no longer hypothetical, and industry and field evidence complement the peer-reviewed literature. Recycled plastic has been incorporated into asphalt at field scale, with published field-performance evaluations of plastic-modified mixes [201] and detailed accounts of the volumetric, workability and mechanical behaviour of waste-plastic-modified mixtures [202]. Plastic-bonded sand blocks produced from waste polyethylene have been manufactured and characterised as a low-cost walling material; this is a non-PET polymer and is cited here only as background evidence [40]. Policy and industry analyses add the market dimension: global plastic waste generation continues to rise while only a small fraction is recycled, so demand-side measures for recycled content are needed to create stable outlets for recyclate [203]."

New references: [201] Lee et al., Proceedings of the International Conference on Road and Airfield Pavement Technology 2023, 135-143; [202] Capitao, Almeida & Picado-Santos, Plastic Waste for Sustainable Asphalt Roads, Elsevier, 2022, 117-144; [203] OECD, Global Plastics Outlook: Policy Scenarios to 2060, 2022.

Comment 1.4

Section 4: The content in this section is divided into too many sub-chapters, and most of the sub-chapters lack in-depth analysis and discussion. Also, many data lack literature support.

Response: We thank the reviewer and have restructured the section accordingly. The twenty third-level subsections have been removed, and Section 4 now contains five subsections, one per material class, each following the same fixed order (incorporation methods, mechanical performance, physical and thermal properties, durability). The detail that previously sat in thin subsections has been either merged into the synthesis or moved to the Supplementary Materials, and each subsection now states explicitly which comparisons are valid and which are not. Study-level values, including dosage bases and test standards, are tabulated in the Supplementary Materials so that the claims in the text are traceable to individual studies.

Revised text (introduction to Section 4): "To keep the synthesis comparable, each subsection covers the same four aspects in a fixed order (incorporation methods, mechanical performance, physical and thermal properties, and durability), and dosage is always stated with the basis on which it was reported. Results obtained with different dosage bases, different PET forms (aggregate, fibre, filler or chemically modified polymer) and different curing regimes are reported separately and are not pooled."

Revised text (Section 4.1.2): "Flexural and splitting tensile strength follow a similar trend, and PET fibres improve post-crack energy absorption and toughness by up to 40% at optimised fibre content [26]. These fibre results are not comparable with the aggregate-replacement results above, because fibre dosage is expressed as a volume fraction of the mix whereas aggregate replacement is expressed relative to the mineral aggregate."

Reviewer 2 Report

Comments and Suggestions for Authors

Thank you for tackling this general topic of recycling PET bottle waste for use in construction in a systematic review. The manuscript includes a look at PET recycling technologies, applications of recycled PET in concrete, asphalt, bricks, insulation and composite panels, along with environmental, health, fire-safety and economic issues. The topic is related to circular construction and waste valorization, however important methodological improvements are required regarding the systematic-review methodology, evidence synthesis, accuracy of reference and interpretation of quantitative evidence.

The following are comments for development:

Lines 11-26: "69 papers were retrieved from Scopus and Web of Science" but later the methodology reports that papers were searched from Scopus, Web of Science, Google Scholar and PubMed Central. Consistently create a description of the database throughout the manuscript.
The statement in lines 21-23, "The optimal PET replacement ratios are "typically around 5-15%", is a mixture of dissimilar materials and replacement bases. There should be no single "optimum" range of concrete aggregate replacements, asphalt binder percentage, fiber volume or brick PET content.
Line 62-79: The research gap is expressed in general terms. Explain the differences between this review and past PET/plastic construction reviews and the additional synthesis that is accomplished beyond the collection of applications.
Lines 81-106: The methodology used in the systematic review is not completely reproducible. Enter search terms, specific search dates, duplicate removal process, number of reviewers screening for duplicates, disagreement resolution, and data-extraction process.
In Figure 1 (page 3), the records are not listed for all four databases reported in the methodology, but only for Scopus and Web of Science. Edit the diagram so that all identification numbers match those found in all the databases that were searched.
Figure 1: The number of articles identified after screening differed from the number of articles included in the final analysis, showing 128 articles after initial screening and 69 articles included in the final analysis in the manuscript, and 798 articles identified, 512 after removing duplicates, 125 reports sought, and 69 finally included in the PRISMA diagram. Ensure all numerical representation of screening process is consistent.
Systematic review methodology: There is no study-quality assessment or risk-of-bias/quality-appraisal procedure reported. A systematic review should provide details of the evaluation of the reliability of the studies included in the review, as there is a significant variation in the methods used in the experiments and in the methods used to incorporate PET.
Lines 99-106: Both original research and systematic reviews may be included. Discuss how evidence was not double counted when review articles were used to inform the primary studies used in the same synthesis.
Mechanical recycling is mentioned as the preferred technique for construction grade PET in line 127–132, which seems to be a generic statement and not a conclusion drawn from a systematic study of all included studies (69). Use numbers to substantiate or temper a statement.
Figure 3 and Table 1: The cost, energy demand and TRL ranges are quantitative comparisons, but midpoint values and ranges are not sufficiently transparent. List the values for the individual sources and describe how the values were normalized and how this was done to compare it to the other data.
Lines 193-212 and Figure 4: Concrete strength synthesis demonstrate smooth lower, midpoint and upper curves, as a function of PET replacement ratio. Describe in detail how these curves are derived from studies reviewed. This should be clearly indicated if they are illustrative, not statistically fitted, relationships.
The use of PET as fine aggregate, coarse aggregate, fiber, and the replacement rates (either by mass or volume) can have a significant impact on the performance of PET concrete. Such combinations should not be mixed.
Lines 229-258: distinction between dry and wet PET modification is not clear enough in the asphalt section. The melting temperature of PET is much higher than the mixing temperatures of conventional asphalt, so it is necessary to clarify and provide evidence for the corresponding processing route that PET is just blended into bitumen at 160–180 °C.
The reported strength increase of Marshall stability of 15-40% at 4-8% PET is a range of increases that may be based upon studies conducted on different forms of PET, dosage definitions, aggregate gradings, binders, and standards used for the Marshall test. Show these conditions in a tabular format and then give a general range.
Table 2 shows the reporting of PET content on various bases that are incompatible from material to material, such as percentage of aggregate replacement, percentage of binder, fiber volume, and weight percentage. Include a dosage basis column and never compare dosage without normalizing.
Environmental benefits (50% energy savings, 79% carbon reduction, 46% GWP reduction) come from different system boundaries and products as shown in lines 362 to 377 and Figure 6. Should not be presented side by side as comparable measures of performance without harmonisation of functional units and of LCA boundaries.
The statement of the toxicity of antimony, phthalates, microplastics, and general plastics is important, but it is not clearly separated from evidence on PET bottles. Ensure consistency in synthesis according to the given scope of PET.
References: Reference list should be fully technically verified. Many of the DOI entries do not match titles of articles or journals cited. For instance, there are references for PET/asphalt, concrete, recycling and environmental which have DOI strings that look like they are from other publications. All DOI, titles, journals, years and page numbers should be verified before publication from the original.
The methodology makes it clear that studies on polymers other than PET were excluded, so why did they include studies on PVC, LDPE, polystyrene, rubber, polypropylene or other polymer materials that are not PET? Make a distinction between background references and the 69 references included in the article and include an additional table listing all included studies.
Lines 528-553: The results of the conclusions include general statements about suitability, the best PET content, economic costs and benefits, and environmental issues. These should be more clearly connected with the type of application and the quality of evidence than expressed as generalizations.

Author Response

We are grateful to the reviewer for a detailed and constructive report. The methodological and evidential points raised have reshaped Section 2 and Section 4, and each is answered below.

Comment 2.1

Lines 11-26: "69 papers were retrieved from Scopus and Web of Science" but later the methodology reports that papers were searched from Scopus, Web of Science, Google Scholar and PubMed Central. Consistently create a description of the database throughout the manuscript.

Response: We thank the reviewer for catching this inconsistency, which was a genuine error. The same four databases are now named in the Abstract, in Section 2.1 and in the caption of Figure 1, and the phrasing is identical in each place.

Revised text (Abstract): "A systematic search of four databases (Scopus, Web of Science, Google Scholar and PubMed Central) returned 798 records; after duplicate removal and two-stage screening, 69 peer-reviewed studies met the eligibility criteria and were included."

Revised text (Section 2): "Four databases were searched: Scopus, Web of Science Core Collection, Google Scholar and PubMed Central."

Comment 2.2

The statement in lines 21-23, "The optimal PET replacement ratios are "typically around 5-15%", is a mixture of dissimilar materials and replacement bases. There should be no single "optimum" range of concrete aggregate replacements, asphalt binder percentage, fiber volume or brick PET content.

Response: We agree entirely, and the single range has been removed from the Abstract. The finding is now reported as four application-specific dosages, each with its base stated, and the manuscript states explicitly that these values are not interchangeable.

Revised text (Abstract): "The findings indicate that PET is suitable for non-structural and semi-structural applications, but that the optimal dosage is application-specific and cannot be reduced to a single range: approximately 5-15% by volume as an aggregate replacement in concrete, 4-8% by weight of binder in asphalt, 25-40% by mass in PET-sand bricks, and 0.5-2% by fibre volume in fibre-reinforced mortar."

Comment 2.3

Line 62-79: The research gap is expressed in general terms. Explain the differences between this review and past PET/plastic construction reviews and the additional synthesis that is accomplished beyond the collection of applications.

Response: We thank the reviewer and have rewritten the gap statement as an explicit comparison in four respects: application scope, polymer specificity, dosage normalisation, and treatment of safety and environmental evidence. The synthesis that this review adds beyond a catalogue of applications is stated directly: a restricted scope to bottle-grade PET, dosage reported on stated bases within each material class, an appraisal of study quality, and environmental and health evidence treated as a primary outcome.

Revised text (Section 1): "Earlier reviews have established that plastic waste can be used in construction, but their scope differs from the present work in four respects. First, several reviews address one application only, for example concrete [12,13] or asphalt [11], and therefore cannot compare performance across material classes. Second, most of them treat plastics generically and do not separate PET from polyvinyl chloride (PVC), low-density polyethylene (LDPE), polypropylene (PP) or polystyrene (PS), although these polymers differ substantially in melting behaviour, stiffness, flammability and additive content [10,13]. Third, dosage is frequently reported on incompatible bases, so optimum values taken from different studies cannot be compared without normalisation. Fourth, safety and environmental risks such as microplastic release, leaching of antimony and phthalate plasticisers, and fire behaviour are seldom integrated into the synthesis, although they are decisive for regulatory approval [14]."

Comment 2.4

Lines 81-106: The methodology used in the systematic review is not completely reproducible. Enter search terms, specific search dates, duplicate removal process, number of reviewers screening for duplicates, disagreement resolution, and data-extraction process.

Response: We thank the reviewer and have rewritten Section 2 so that each of these elements is reported. The database-specific strategies are given, with the field codes used in Scopus and Web of Science and the core Boolean string; duplicate removal is described; the number of screeners and the mechanism for resolving disagreement are stated; and the data-extraction form is listed field by field.

Revised text (Section 2): "In Scopus the query was applied to title, abstract and keywords (TITLE-ABS-KEY) and in Web of Science to topic fields (TS); Google Scholar and PubMed Central were searched with the same terms adapted to their query syntax. The core string was: ("polyethylene terephthalate" OR "PET" OR "poly(ethylene terephthalate)") AND (recycl* OR reuse OR valoris* OR valoriz* OR upcycl*) AND ("building material*" OR construction OR concrete OR cement* OR mortar OR asphalt OR bitumen OR brick* OR insulation OR panel* OR composite*)."

Revised text (Section 2): "Two reviewers ([AUTHOR_INPUT: initials]) independently screened titles and abstracts against the eligibility criteria, and then screened the full texts of the retained records. Disagreements were resolved by discussion and, where agreement could not be reached, by a third reviewer ([AUTHOR_INPUT: initials]). Agreement before resolution was quantified with Cohen's kappa ([AUTHOR_INPUT: value])."

Revised text (Section 2): "The following fields were extracted: PET waste source and form (flake, granule, powder, fibre, chemically modified or depolymerised); matrix material; dosage and the basis on which dosage was reported; specimen preparation, curing and conditioning; test standard and specimen geometry; and all reported performance outcomes."

Comment 2.5

In Figure 1 (page 3), the records are not listed for all four databases reported in the methodology, but only for Scopus and Web of Science. Edit the diagram so that all identification numbers match those found in all the databases that were searched.

Response: We thank the reviewer and have corrected the caption so that the diagram is described explicitly as reporting all four databases. The identification, screening and inclusion numbers in the figure are now stated in the caption, so that the reader can verify that the figure and the text agree. The diagram itself is being redrawn so that the identification block is broken down by database.

Revised caption (Figure 1): "Records are reported separately for each of the four databases searched (Scopus, Web of Science, Google Scholar and PubMed Central). In total 798 records were identified, 512 remained after duplicate removal, 128 passed title and abstract screening, 125 full texts were sought and assessed, and 69 studies were included in the synthesis."

Comment 2.6

Figure 1: The number of articles identified after screening differed from the number of articles included in the final analysis, showing 128 articles after initial screening and 69 articles included in the final analysis in the manuscript, and 798 articles identified, 512 after removing duplicates, 125 reports sought, and 69 finally included in the PRISMA diagram. Ensure all numerical representation of screening process is consistent.

Response: We thank the reviewer. The text and the figure are now reconciled to a single set of numbers, which are also reported in the methodology. The previous sentence stating that 128 articles were identified "after title/abstract and full-text screening" has been removed, since it conflated two different stages and did not match the diagram.

Revised text (Section 2): "The searches identified 798 records. After duplicate removal, 512 records remained; 128 records passed title and abstract screening; full texts were sought for 125 of these and assessed for eligibility; and 69 studies met all criteria and were included in the synthesis."

Comment 2.7

Systematic review methodology: There is no study-quality assessment or risk-of-bias/quality-appraisal procedure reported. A systematic review should provide details of the evaluation of the reliability of the studies included in the review, as there is a significant variation in the methods used in the experiments and in the methods used to incorporate PET.

Response: We agree that this was a significant omission and thank the reviewer for identifying it. A quality and risk-of-bias appraisal has been added. Because the included studies are experimental and largely non-randomised, a purpose-built six-item instrument was used rather than a randomised-trial tool, and studies classified as high risk of bias are now excluded from the quantitative ranges reported in Section 4.

Added text (Section 2): "Because the included studies are experimental and mostly non-randomised, a purpose-built appraisal tool was used rather than a randomised-trial instrument. Each study was scored on six items, each rated 0 to 2: clarity of the PET feedstock and dosage basis; reporting of mix proportions; reporting of specimen preparation and curing; use of a recognised test standard; reporting of replicate numbers and dispersion; and comparison against a PET-free reference mix. The total score (0-12) was used to classify studies as low (10-12), moderate (7-9) or high (less than or equal to 6) risk of bias. Studies classified as high risk of bias were retained for qualitative description only and were excluded from the quantitative ranges reported in Section 4."

Comment 2.8

Lines 99-106: Both original research and systematic reviews may be included. Discuss how evidence was not double counted when review articles were used to inform the primary studies used in the same synthesis.

Response: We thank the reviewer for this point, which is now stated explicitly in the methodology.

Added text (Section 2): "Where both a review article and one of its primary studies met the eligibility criteria, the primary study was used for data extraction and the review was retained only for background and for identifying further primary studies. No dataset therefore contributed twice to the syntheses in Section 4, and only primary studies were used to derive the reported dosage and performance ranges."

Comment 2.9

Mechanical recycling is mentioned as the preferred technique for construction grade PET in line 127-132, which seems to be a generic statement and not a conclusion drawn from a systematic study of all included studies (69). Use numbers to substantiate or temper a statement.

Response: We agree and have tempered the claim. The statement is now tied to the feedstock route reported by the included studies, and the manuscript states the limitation that property retention after repeated processing is rarely quantified in those studies, so the claim is presented as evidence-based for the included set rather than as a general rule.

Revised text (Section 3.1): "Of the 69 included studies, [AUTHOR_INPUT: number] reported the feedstock preparation route; mechanical routes dominated, while chemical and enzymatic routes appeared mainly in technology-assessment studies rather than in building-material performance studies. Mechanical recycling is therefore best described as the dominant and most cost-effective route for construction-grade rPET on the available evidence rather than as universally preferred, because property retention after repeated thermal-mechanical cycles is rarely quantified in the included studies and the evidence base is unevenly distributed across applications."

Comment 2.10

Figure 3 and Table 1: The cost, energy demand and TRL ranges are quantitative comparisons, but midpoint values and ranges are not sufficiently transparent. List the values for the individual sources and describe how the values were normalized and how this was done to compare it to the other data.

Response: We thank the reviewer. The captions now define how midpoints were obtained and where the underlying values are reported, and a supplementary table lists the per-source values, the unit conversions and the normalisation applied. The captions also state plainly that the panels are indicative and are not normalised to a common functional unit.

Revised caption (Figure 3): "Bars show the midpoint of the ranges reported by the individual sources listed in Table 1, where the midpoint is the arithmetic mean of the reported lower and upper bound. Unit conversions and the normalisation procedure are described in Supplementary Table S3, together with the per-source values."

Comment 2.11

Lines 193-212 and Figure 4: Concrete strength synthesis demonstrate smooth lower, midpoint and upper curves, as a function of PET replacement ratio. Describe in detail how these curves are derived from studies reviewed. This should be clearly indicated if they are illustrative, not statistically fitted, relationships.

Response: We thank the reviewer for pressing on this point. The curves are not statistical fits, and the figure caption now says so explicitly and describes how they were drawn. The underlying study-level values and the derivation of the envelopes are reported in the Supplementary Materials. We have also separated the data by PET form, because aggregate replacement, fibre addition and filler addition are not comparable.

Revised caption (Figure 4): "The curves are illustrative trend envelopes drawn by hand through the midpoints of the ranges reported by the included studies; they are not regression fits and no statistical model was estimated. Individual studies are plotted separately by PET form (aggregate replacement, fibre addition and filler addition), because the dosage bases are not interchangeable, and the shaded band spans the reported range. The source values and the derivation of the envelopes are given in Supplementary Table S4."

Comment 2.12

The use of PET as fine aggregate, coarse aggregate, fiber, and the replacement rates (either by mass or volume) can have a significant impact on the performance of PET concrete. Such combinations should not be mixed.

Response: We agree, and the three forms are now treated separately throughout Section 4.1, with an explicit statement of the basis on which each is reported.

Added text (Section 4.1.1): "These three forms are reported on different bases: aggregate replacement is expressed relative to the mass or volume of mineral aggregate, fibre addition as a fibre volume fraction, and filler addition relative to the mass of cement. The three groups are therefore treated separately throughout this review."

Comment 2.13

Lines 229-258: distinction between dry and wet PET modification is not clear enough in the asphalt section. The melting temperature of PET is much higher than the mixing temperatures of conventional asphalt, so it is necessary to clarify and provide evidence for the corresponding processing route that PET is just blended into bitumen at 160-180 degrees C.

Response: We thank the reviewer for identifying what was an internal contradiction in the manuscript. The asphalt subsection has been rewritten to separate the dry and wet routes, and the melting-point problem is now stated explicitly: blending PET into bitumen at 160-180 degrees C is only possible if the PET has been pre-melted, depolymerised or introduced as a low-melting PET-based additive. The manuscript now states that studies reporting wet modification without describing how the PET entered the binder are not reproducible as described, and that this is one reason for the wide variation in reported rheological gains.

Revised text (Section 4.2.1): "In the dry process, PET granules or flakes (typically 1-5 mm) partially replace fine aggregate and remain a discrete particulate phase; the PET is not required to melt, and the mixture is produced at conventional temperatures of about 160-180 degrees C [34]. In the wet process, PET is first incorporated into the binder, either by high-shear blending with bitumen at 160-180 degrees C, which can only succeed if the PET is pre-melted, chemically depolymerised or introduced as a low-melting PET-based additive, because the melting range of bottle-grade PET (approximately 250-260 degrees C) lies well above normal bitumen mixing temperatures [33,35]."

Comment 2.14

The reported strength increase of Marshall stability of 15-40% at 4-8% PET is a range of increases that may be based upon studies conducted on different forms of PET, dosage definitions, aggregate gradings, binders, and standards used for the Marshall test. Show these conditions in a tabular format and then give a general range.

Response: We agree. The conditions reported by each study are now tabulated in the Supplementary Materials, and the text states that the 15-40% range is an envelope across those conditions rather than a transferable single result.

Revised text (Section 4.2.2): "Reported Marshall stability increases of 15-40% at PET contents of 4-8% are not obtained under a single set of conditions: the studies differ in PET form (flake, granule or fibre), dosage basis, aggregate gradation, binder grade and the Marshall test standard applied. The conditions reported by each study are tabulated in Supplementary Table S5, and the range above should be read as an envelope across those conditions rather than as a single transferable result."

Comment 2.15

Table 2 shows the reporting of PET content on various bases that are incompatible from material to material, such as percentage of aggregate replacement, percentage of binder, fiber volume, and weight percentage. Include a dosage basis column and never compare dosage without normalizing.

Response: We thank the reviewer. The mixed "PET content" column has been replaced by an explicit dosage column in which the basis is stated for every row, and a note under the table states that the values are not comparable across rows. The methodology now also states that dosages reported on different bases were never pooled and were converted within a material class only where the conversion was reported by the original authors.

Revised Table 2 column header and values: "PET dosage (basis)"; "5-15% by volume of fine or coarse aggregate"; "4-8% by mass of bitumen binder"; "25-40% by mass of dry mix"; "0.5-2% fibre volume fraction"; "80-100% by mass of product".

Added note (Table 2): "Dosages in Table 2 are expressed on different bases and are therefore not directly comparable between rows. Aggregate replacement, binder mass fraction, fibre volume fraction and product mass fraction are reported separately, and no normalisation across bases has been attempted."

Comment 2.16

Environmental benefits (50% energy savings, 79% carbon reduction, 46% GWP reduction) come from different system boundaries and products as shown in lines 362 to 377 and Figure 6. Should not be presented side by side as comparable measures of performance without harmonisation of functional units and of LCA boundaries.

Response: We agree, and the paragraph has been rewritten so that each figure is attributed to its own functional unit, system boundary and product. The manuscript now states explicitly that the three values are not directly comparable as published and that a harmonised comparison is not possible with the current literature. The figure caption carries the same caveat.

Revised text (Section 5.1): "These figures must be read with care. They come from studies with different functional units (1 kg of polymer, 1 kg of binder and 1 m2 of insulation panel), different system boundaries (cradle-to-gate and cradle-to-grave), different allocation choices and different products. They are therefore not directly comparable and are reported here as separate study-specific results rather than as a common performance metric."

Revised caption (Figure 6): "Each bar originates from a different study, product and system boundary. The values are not normalised to a common functional unit and must not be read as directly comparable."

Comment 2.17

The statement of the toxicity of antimony, phthalates, microplastics, and general plastics is important, but it is not clearly separated from evidence on PET bottles. Ensure consistency in synthesis according to the given scope of PET.

Response: We thank the reviewer and have added an explicit scope statement. Evidence relating to other polymers is now identified as such and is included only where it is directly transferable. A factual error in this subsection has also been corrected: the original text stated that chronic and cumulative effects are fully understood, whereas the cited source indicates that they are not.

Added text (Section 5.2): "Unless stated otherwise, the evidence in this section concerns PET; mechanisms reported for other polymers are included only where they are directly transferable, and they are identified as such."

Corrected text (Section 5.2.2): "Ecotoxicological research remains incomplete: although the acute toxicity of PET leachates to aquatic organisms appears to be low, chronic and cumulative effects are not yet fully understood [55]."

Comment 2.18

References: Reference list should be fully technically verified. Many of the DOI entries do not match titles of articles or journals cited. For instance, there are references for PET/asphalt, concrete, recycling and environmental which have DOI strings that look like they are from other publications. All DOI, titles, journals, years and page numbers should be verified before publication from the original.

Response: We thank the reviewer for this important observation and confirm that it was correct. We have re-verified all 200 references programmatically against Crossref and, for the affected entries, against the publisher record. Every DOI was resolved and the returned title, journal, volume, page and year were compared with the cited reference. The outcome is summarised below and is listed reference by reference in the reference verification log supplied with this revision.

• 15 references had a DOI that resolved to a different article; each DOI has been replaced with the verified DOI.
• 11 references had no DOI; a verified DOI has been added.
• 24 DOIs could not be matched to the cited work by any route and have been removed rather than left incorrect; a further 2 entries are flagged in the log for confirmation against the original sources.
• 138 further entries were not cited anywhere in the manuscript; these have been removed from the reference list, which also resolved the duplicate entries detected during verification.
• A machine-readable log (reference_verification_log.csv) accompanies this revision and gives, for each reference, the outcome, the DOI as submitted, the verified DOI and the authoritative title.

Note: where a reference could not be verified, we have not substituted a plausible-looking DOI. The log identifies these entries explicitly so that they can be confirmed from the original sources before publication.

Comment 2.19

The methodology makes it clear that studies on polymers other than PET were excluded, so why did they include studies on PVC, LDPE, polystyrene, rubber, polypropylene or other polymer materials that are not PET? Make a distinction between background references and the 69 references included in the article and include an additional table listing all included studies.

Response: We thank the reviewer for this point. The methodology now states that studies addressing other polymers were retained only as background and that they are identified as such, a note at the head of the reference list explains the marking convention, and a supplementary table has been added to record the characteristics of the included studies (Supplementary Table S1). In the Introduction, the one non-PET case that is cited is now labelled explicitly as background evidence.

Added text (Section 2): "Studies addressing other polymers were retained only as background references and are identified as such in the reference list."

Added note (References): "References marked with a dagger are background sources, including studies on polymers other than PET and policy or industry reports; they were not part of the 69 included studies. The characteristics of all 69 included studies are listed in Supplementary Table S1."

Comment 2.20

Lines 528-553: The results of the conclusions include general statements about suitability, the best PET content, economic costs and benefits, and environmental issues. These should be more clearly connected with the type of application and the quality of evidence than expressed as generalizations.

Response: We agree and have rewritten Section 8. Each conclusion is now stated for a defined application and is qualified by the strength of the supporting evidence, and the environmental figures are reported with their system boundaries attached rather than as general performance claims.

Revised text (Section 8): "The evidence reviewed here supports four conclusions, each of which is stated with the application and the quality of evidence to which it applies."

Revised text (Section 8): "The evidence for bricks, however, rests on a small number of studies with different sand contents and compaction procedures, so the reported 5-25 MPa range should be treated as indicative rather than as a specification."

Revised text (Section 8): "Until harmonised life-cycle assessments are available, these figures should be quoted with their system boundary attached rather than as general performance claims."

Reviewer 3 Report

Comments and Suggestions for Authors

This manuscript reviews the recycling of polyethylene terephthalate (PET) bottle waste in building materials. The topic is relevant to circular construction and has practical potential. However, major concerns remain regarding the review methodology, consistency of the evidence base, quantitative synthesis, and reference accuracy. Therefore, the comments are as follows.

• Section 1.2: The claimed gap is not supported by a systematic comparison with existing reviews, so the novelty of the present review remains unclear.
• Section 2: The search description lacks database-specific strategies and search dates, and the database list conflicts with the Abstract, limiting reproducibility.
• Which 69 studies were included in the review, particularly given that the reference list contains 200 items as well as pre-2000 and non-PET studies that appear to conflict with the eligibility criteria?
• The synthesis combines different PET forms, replacement bases, mix designs, curing conditions, and test methods, but the general optimum replacement ranges are not adequately justified across these heterogeneous studies.
• What are the original data sources and calculation methods for Figures 3-7, and how were the displayed values derived from the reviewed studies?
• Section 8: The conclusions generalize across materially different applications without sufficient study-level support for the stated optimum ratios and broad environmental benefits.
• The reference list contains numerous DOI entries that appear inconsistent with the cited titles or journals, together with duplicated or apparently irrelevant references, which undermines the traceability of the evidence.

Overall, the topic has practical value, but the review methodology, evidence synthesis, and reference accuracy require substantial improvement. Major revision is recommended.

Author Response

We thank the reviewer for a concise and well-targeted report. The methodological, evidential and referencing concerns raised are closely aligned with those of the other reviewers, and each is answered fully below, with the corresponding manuscript locations identified.

Comment 3.1

Section 1.2: The claimed gap is not supported by a systematic comparison with existing reviews, so the novelty of the present review remains unclear.

Response: We thank the reviewer. The gap statement has been rewritten as an explicit comparison with existing reviews, and it is now expressed in four checkable respects rather than as a general claim of scarcity. The former subsection heading has also been removed, so the argument now runs continuously within Section 1.

Revised text (Section 1): "Earlier reviews have established that plastic waste can be used in construction, but their scope differs from the present work in four respects. First, several reviews address one application only, for example concrete [12,13] or asphalt [11], and therefore cannot compare performance across material classes. Second, most of them treat plastics generically and do not separate PET from polyvinyl chloride (PVC), low-density polyethylene (LDPE), polypropylene (PP) or polystyrene (PS), although these polymers differ substantially in melting behaviour, stiffness, flammability and additive content [10,13]. Third, dosage is frequently reported on incompatible bases, so optimum values taken from different studies cannot be compared without normalisation. Fourth, safety and environmental risks such as microplastic release, leaching of antimony and phthalate plasticisers, and fire behaviour are seldom integrated into the synthesis, although they are decisive for regulatory approval [14]."

Comment 3.2

Section 2: The search description lacks database-specific strategies and search dates, and the database list conflicts with the Abstract, limiting reproducibility.

Response: We agree and thank the reviewer. The database list is now identical in the Abstract, Section 2 and Figure 1. Database-specific field codes and the full Boolean string are reported, the search window is stated, and the screening, disagreement-resolution and data-extraction procedures are described so that the process can be repeated.

Revised text (Section 2): "The searches were run on [AUTHOR_INPUT: date on which the searches were last executed] and covered the period from 1 January 2000 to 31 May 2025. Database-specific strategies were used rather than one generic string. In Scopus the query was applied to title, abstract and keywords (TITLE-ABS-KEY) and in Web of Science to topic fields (TS); Google Scholar and PubMed Central were searched with the same terms adapted to their query syntax."

Comment 3.3

Which 69 studies were included in the review, particularly given that the reference list contains 200 items as well as pre-2000 and non-PET studies that appear to conflict with the eligibility criteria?

Response: We thank the reviewer for identifying a problem that was more serious than it appeared. On re-checking the reference list we found that only 62 of the 200 entries were cited anywhere in the manuscript; the remaining 138 entries were never cited and have now been removed. Of the 62 cited entries, a small number are background sources covering plastics generically, a non-PET polymer or construction and demolition waste; these are now marked with a dagger at the head of the reference list and are identified as background rather than as included studies. The eligibility window of January 2000 to May 2025 applies to the included studies; background sources, including policy and industry reports, are not subject to it and are labelled as such.

Added note (References): "References are numbered in order of first citation. Entries preceded by a dagger are background sources, including studies on polymers other than PET and policy or industry reports; they are not part of the 69 included studies. The characteristics of the 69 included studies are listed in Supplementary Table S1."

Added text (Section 2): "Studies addressing other polymers were retained only as background references and are identified as such in the reference list."

Comment 3.4

The synthesis combines different PET forms, replacement bases, mix designs, curing conditions, and test methods, but the general optimum replacement ranges are not adequately justified across these heterogeneous studies.

Response: We agree, and the synthesis has been restructured so that these sources of heterogeneity are handled explicitly rather than averaged over. Dosage is now always stated with its basis, PET forms are reported separately within each material class, and dosages expressed on different bases are never pooled. The methodology also states the extraction fields that capture mix design, curing, conditioning and test standard, and study quality is appraised so that high-risk studies are excluded from the quantitative ranges.

Revised text (Section 4 introduction): "Results obtained with different dosage bases, different PET forms (aggregate, fibre, filler or chemically modified polymer) and different curing regimes are reported separately and are not pooled."

Revised text (Section 4.1.1): "These three forms are reported on different bases: aggregate replacement is expressed relative to the mass or volume of mineral aggregate, fibre addition as a fibre volume fraction, and filler addition relative to the mass of cement."

Comment 3.5

What are the original data sources and calculation methods for Figures 3-7, and how were the displayed values derived from the reviewed studies?

Response: We thank the reviewer. The captions now state, for each figure, the derivation of the plotted values, the units, the sources and the limitations of the comparison. Where a figure aggregates values from several sources, the per-source values, unit conversions and normalisation steps are given in the Supplementary Materials, and where the plotted relationship is illustrative rather than statistical, the caption says so explicitly.

Revised caption (Figure 4): "The curves are illustrative trend envelopes drawn by hand through the midpoints of the ranges reported by the included studies; they are not regression fits and no statistical model was estimated. ... The source values and the derivation of the envelopes are given in Supplementary Table S4."

Revised caption (Figure 6): "Each bar originates from a different study, product and system boundary. The values are not normalised to a common functional unit and must not be read as directly comparable."

Comment 3.6

Section 8: The conclusions generalize across materially different applications without sufficient study-level support for the stated optimum ratios and broad environmental benefits.

Response: We agree, and Section 8 has been rewritten. Each conclusion is now attached to a named application and qualified by the strength of the supporting evidence, and the environmental figures are reported with their system boundaries rather than as general claims.

Revised text (Section 8): "The evidence reviewed here supports four conclusions, each of which is stated with the application and the quality of evidence to which it applies."

Revised text (Section 8): "The evidence for bricks, however, rests on a small number of studies with different sand contents and compaction procedures, so the reported 5-25 MPa range should be treated as indicative rather than as a specification."

Comment 3.7

The reference list contains numerous DOI entries that appear inconsistent with the cited titles or journals, together with duplicated or apparently irrelevant references, which undermines the traceability of the evidence.

Response: We thank the reviewer; the concern was well founded and has been addressed by a complete re-verification of the reference list. Every DOI was resolved and the returned title, journal, volume, page and year were compared with the cited entry. Duplicate entries were identified, and because all of them fell in the uncited portion of the list they were removed with the uncited entries. The remaining list contains only cited sources.

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

Suitable modifications were done by authors.

Author Response

Comment 1

Suitable modifications were done by authors.

Response: We thank the reviewer for the assessment. The report does not identify an outstanding item, so no change to the manuscript was required in response to it. We confirmed that position before resubmission by re-reading the manuscript against the previous round of comments and checking that the revisions described in that round are still present and unaltered in substance. As a closing check we also re-read the manuscript for consistency between the Abstract, the Methods and the figures and tables. The study design, the evidence base and the conclusions are unchanged from the version assessed.

Reviewer 2 Report

Comments and Suggestions for Authors

Thanks for addressing the comments
the abstract says 63 studies while the methodology says 69!

Author Response

Comment 1

The abstract says 63 studies while the methodology says 69!

Response: We agree, and we are grateful to the reviewer for identifying this. The count in the Abstract was the incorrect one. The study selection described in Section 2 yields 69 included studies, and the Abstract has been corrected from 63 to 69.

The selection outcome reported in Section 2 reconciles at every step. The searches identified 798 records. After duplicate removal, 512 records remained and were screened on title and abstract, of which 384 were excluded. Full texts were sought for the remaining 128 records, three of which could not be retrieved, leaving 125 reports assessed for eligibility. Of these, 56 were excluded at the full-text stage, comprising full text not available (12), duplicate dataset or companion paper (18), no PET-specific data (15) and off-topic application (11). The remaining 69 studies met all criteria and were included in the review, and 28 of them reported quantitative performance data on a dosage basis that could be extracted.

The Abstract now states 69 studies, and it continues to state that 28 of the included studies reported the quantitative data from which the reported ranges were derived, so the two counts agree with Section 2, with Table 2 and with Supplementary Table S1.

We also checked the remaining counts given in the Abstract against the Methods and the results sections. The four databases searched, the three recycling routes compared, the five product categories reviewed, the 798 records identified and the 28 studies contributing extractable data agree throughout, and no further inconsistency was found.

Revised manuscript text (Abstract, Section 1). Previous wording,

“...returned 798 records; after duplicate removal and two-stage screening, 63 studies met the eligibility criteria, of which 28 reported the quantitative performance data...”

Revised wording, “A systematic search of four databases (Scopus, Web of Science Core Collection, Google Scholar and PubMed Central) returned 798 records; after duplicate removal and two-stage screening, 69 studies met the eligibility criteria, of which 28 reported the quantitative performance data that underpin the ranges extracted here.”

Reviewer 3 Report

Comments and Suggestions for Authors

The authors have revised the manuscript as per the comments. After re-reviewing the manuscript, I still suggest the authors making minor revisions to the figures and conclusions. For the figures they should be double checked to meet the academic standard. In addition, the conclusions should be presented point by point free of citations. After the adoption of comments for revisions, I agree to have this manuscript accepted for publication.

Author Response

Comment 1

The authors have revised the manuscript as per the comments. After re-reviewing the manuscript, I still suggest the authors making minor revisions to the figures and conclusions. For the figures they should be double checked to meet the academic standard. In addition, the conclusions should be presented point by point free of citations. After the adoption of comments for revisions, I agree to have this manuscript accepted for publication.

Figures

Response: We thank the reviewer for this instruction. Every figure, table and caption has been re-checked against the journal's presentation conventions. The check identified two formatting matters and one placement inconsistency, which we have corrected, and it confirmed the construction of the figures themselves.

Caption styles. The captions of Figures 1 and 3 and of Tables 1, 2 and 3 had been set in the body-text style rather than in the caption styles of the template. All seven figure captions now use the figure-caption style and all three table captions now use the table-caption style, so that the caption font, size, indentation and spacing follow the journal template.

Caption punctuation. The caption of Figure 1 did not end with a full stop. It now reads “Figure 1. PRISMA 2020 flow diagram of the study selection process.”

Figure placement. Figures 1 and 2 had been set with right-hand alignment while Figures 3 to 7 were centred. Both are now centred, so that all seven figures share a single horizontal placement.

We also verified how the figures are constructed. Figures 3 to 7 are native, editable chart objects rather than images, twelve charts in total. Each chart is 6.90 in wide, and the panels within each multi-panel figure are identical in size, so those figures are internally uniform. Chart titles, axis labels, series names and the legend are editable in the source file and remain legible at print size. Figures 1 and 2 are the two raster figures. Both were inspected at print size. Figure 1 is supplied at 1000 by 1000 pixels and Figure 2 at 1447 by 1208 pixels, and both render completely and legibly, with no clipping.

All of these changes are shown as tracked changes in the revised manuscript.