Semi-quantitative Computational Analysis of Plastic Additives in a FLOPP-E and SLOPP-E Database Subset

Semi-quantitative Computational Analysis of Plastic Additives in a FLOPP-E and SLOPP-E Database Subset | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Semi-quantitative Computational Analysis of Plastic Additives in a FLOPP-E and SLOPP-E Database Subset Wesley Allen Williams, Shyam Aravamudhan This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-5334015/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 05 Mar, 2025 Read the published version in Microplastics and Nanoplastics → Version 1 posted 9 You are reading this latest preprint version Abstract Microplastic (MPL) abundance in the environment and the biosphere is a grave problem that is confounded by many aspects, one vital aspect being the characterization of their heterogeneous matrix. Currently, spectroscopy, chromatography, and soxhelation aid in this matter. However, many of these techniques are time consuming for MPL characterization, which can include a large number of particles. Therefore, we propose a facile “Additive Analysis” algorithm that can provide the top ten matches for additives for an MPL. For our first trial, we used 2 MPL entries, from FLOPP-E (C2. Blue Fiber) and SLOPP-E (Polyester 12. Red Fiber), as a continuation of our previous work. For our second trial, we extended the use of the algorithm to a semi-randomly selected subset of MPL samples from FLOPP-E and SLOPP-E based on choosing 1 sample of each color for each polymer. Both trials’ reference used an in-lab digitization of the Hummel database for Fourier-transform Infrared (FTIR) spectroscopy and an open-source Raman spectroscopy database from Nava. We determined that the “C2. Blue Fiber” contains amounts of a metal-free phthalocyanine, potentially indicating the presence of degradation in context to the controls (t 10,.05 : .4879, p: .6387). For “Polyester 12. Blue Fiber,” we determined a high likelihood of significant amounts of quinone and azo-family colorants in the sample, negating a previous hypothesis of pyrrole presence (W: 0, p: .036364). For the second trial, 49/56 and 27/40 hits were generated out of the randomly selected samples, with a vast majority possessing hits (matching the color of the sample) within our most scrutinizing tolerance of 5 1/cm (77.6%/74.07%), respectively. For the FTIR portion, the top 3 IDs from tolerances of 5, 10, and 15 1/cm were benzenesulfonohydrazide (1st and 2nd Hit), titanium dioxide (4th Hit), and barium permanganate/barium sulfate (6th Hit). For the Raman portion, the top 3 IDs from tolerances of 5, 10, and 15 1/cm were PR210 (azo derivative – 2nd Hit), PB25 (azo derivative – 2nd Hit), and muscovite (mineral – 1st Hit). Lastly, the distribution for these hits appears to identify organic colorants (FTIR) and azo-derivative colorants (Raman) most dominantly. Our discussion concludes with the potential toxicological impacts of these top 6 IDs. Figures Figure 1 Figure 2 Figure 3 Figure 4 1. INTRODUCTION MPLs are microparticles derived from industry (primary) or derived from environmental breakdown (secondary). The former type of MPLs are typically utilized as “microbeads found in personal care products” [ 1 ] for abrasive effects like shower gels [ 2 ] or are found in synthetic textiles (laundered nylon, polyester, etc.)[ 1 ][ 2 ], tires (via erosion from driving motor vehicles [ 2 ]), and paint-based markings and resins (presumably acrylics from paint primers)[ 3 ]. Some primary MPLs are made to adapt the material properties of other plastic products [ 3 ]. Secondary MPLs originate from environmental forces such as abrasion and photo-oxidation via chain scission events [ 4 ], like the Norrish reaction [ 5 ][ 6 ] which would vary as a function of the body of water’s depth [ 7 ]; thermal oxidation (primarily via chain unfolding [ 8 ] which would vary environmentally based on the rate of the heating and surrounding medium [ 9 ]); microbial digestion [ 10 ]; weathering from water [ 11 ]; and pH (secondary effect: degradation in water is thought to acidify ocean environments [ 12 ] while other reports [ 13 ] specify alkalinization in a terrestrial environment due to additives within the MPLs). As an aside, pH could, in theory, affect degradation by way of preferentially selecting alcohols for oxonium ion formation (leaving group). Many of these particles are inducted into the hydrosphere via various anthropogenic activities (waste, storm runoff, and maritime activities [ 14 ]) where they cause potent toxicity to marine biodiversity [ 15 ] as well as higher-order consumers, like Megaptera [ 16 ] or human beings [ 17 ][ 18 ]. It is obvious that the plastic polymer type may possess interactions in a living system, but are the plastic additives causing an effect in tandem? Indeed, some reports have indicated or theorized that the presence of organic molecules, like plasticizers, in the human body is a causative agent for toxicity. One such report indicated the presence of copper phthalocyanine and hematite, a common blue and red/black pigment in plastic [ 19 ], in human kidneys [ 20 ]! There are many forms of additives but phthalate esters, bisphenol A (BPA), and tributyltin have been selected to encapsulate the significance of this study. Chiefly, plasticizers, like phthalate esters, have been shown [ 21 ] to hypertrophy thyroid cells, increase thyroid activity, and disrupt the hypothalamic-pituitary-thyroid axis (responsible for response to stress and regulation of metabolism). For the pituitary system, in murine models, the dysregulation of the hypothalamic-pituitary-gonadal (HPG) axis is seen affecting the release of gonadotropin-releasing hormone, follicle-stimulating hormone, and luteinizing hormone (impacts spermatogenesis and ovulation) [ 22 ]. In terms of the reproductive system, in murine models, oxidative stress, altered sperm physiology, and anti-androgenic effects are seen [ 21 ] with phthalate exposure. Ovarian steroidogenesis effects and reduced luteinizing hormone are apparent [ 21 ], as well. Generally, phthalate esters have been shown to impact prenatal and postnatal development [ 23 ]. For BPA, alteration in T4 levels, oxidative damage, and antagonism have been shown [ 21 ] in murine models while also affecting the blood-testis-barrier (BTB), significantly disrupting the ability to select against certain small molecules entering the glands. Moreover, spermatozoan DNA damage is induced by BPA, as well while in the female reproductive system, progesterone and oestradiol are inhibited. In the pituitary region, hypothalamic inflammation can occur [ 21 ]. In general, in the past, BPA has been shown to increase childhood obesity and cardiovascular disease incidence [ 23 ]. Lastly, for Tributyltin, thyroid follicle reduction and steroidogenic enzyme activity are impacted, leading to impaired testosterone production and defective spermatozoa [ 21 ]. Many of these additives may potentially impact metabolic activity and the fecundity of mammalian species, including human beings. In general, tin exposure could cause cancer, gastrointestinal (GI) issues, and potential clastogenesis (breakage, loss, or rearrangement of chromosomes) [ 23 ]. Current efforts to characterize plastic additives within pristine and environmental MPLs in the literature range from spectroscopic and chromatographic techniques, sometimes used with rigorous extraction protocols. One such protocol is the Soxhlet extraction process (soxhelation). Soxhelation involves using the special Soxhlet apparatus, which helps condense evaporated solvents and siphon them back to the round bottom flask. This constant recycling minimizes the amount of solvent evaporated out of the apparatus. At the same time, the central vessel allows for the collection of the small molecule or substance under investigation after recondensation. The concentrated extract is further processed for spectroscopic and chromatographic techniques. A technical report [ 24 ] from ThermoFisher details the 6-hour soxhelation of PVC resulting in an essentially 100% recovery (actually > 100%) of DOA (dioctyl adipate - plasticizer), TOP (trioctyl phosphate – flame retardant/plasticizer ) , DOP (dioctyl phthalate – plasticizer), and TOTM ( tris- (2-ethylhexyl)trimellitates – PVC plasticizer) (confirmed via gas chromatography). Other reports have used characterization techniques along with simpler extraction of whole particles, such as oleophilic extraction, in the next report. One such report [ 25 ], utilizing µ-FTIR characterization, determined the constituents of extracted MPLs and microparticles to be an array of polymers and additives. More specifically, they surmise that the incidence of particles primarily owing to plastic additives (UV stabilizers, lubricants, fillers, antioxidants, and colorants, to name a few) may correlate with MPL abundance in the environment. Lastly, for chromatographic techniques, a report [ 26 ] using pyrolysis gas chromatography/mass spectrometry determined the presence of various plasticizers (e.g., phthalates, benzaldehyde, and 2,4-di- tert -butylphenol), titania, barium, zinc, and sulfur. In addition to the additives presented, thus far, plastic polymers have a plethora of additive types that may cause some effect due to their molecular structure like plasticizers, flame retardants, antistatics, antioxidants, surfactants, vulcanizing agents, soaps, crosslinkers, peptizers, foaming agents, lubricants, fillers, curing agents, biocides, metal deactivators, antiozonants, anti-blocking agents, and colorants [ 19 ]. Potentially, computational methods may quicken the identification of these additives in MP matrices. Currently, a common metric of spectral matching is the hit-quality index (HQI). HQI is the dot product of two y-vectors (unknown/reference), squared, divided by the dot product of the squared vectors. This essentially measures covariance between the two samples, meaning perfectly parallel vectors will reach a perfect value of one, and a completely independent set of vectors will reach the lowest value of 0. This has been used [ 27 ] in a novel 2D-HQI technique using both Raman and FTIR HQIs of environmentally derived MPLs. Other reports [ 28 ] utilize aspects from linear algebra, like in one report where significant factors were determined by SVD, spanned by the minima and maxima of each variable in spectra, and the non-negative factor analysis is initiated which looks at the contribution of the factor and the sample. Iterations are run at a set limit to overwrite negative values. Essentially, the user will be able to see a relationship between the samples and variables. However, one caveat is that irrelevant sections must be trimmed as a pre-processing technique. According to Hummel [ 19 ], various other techniques apply a set tolerance to an individual band that arises as a match; thus, different bands are given no calculation. In other cases, the total number of bands in a spectrum that matches “pinged” is used as a metric to control for the match contribution or sometimes used with an arbitrary scaling factor. For sample and reference matching, “Lowry-Huppler” algorithms contain a four-fold whole-spectrum search which is the summation of: 1. The magnitude of the dot product of the spectrum and reference vectors, 2. The previous value, squared, 3 and 4. The same as 1 and 2 except that the tolerance is applied, presumably, scaling the vectors. All vectors must be normalized absorbance values. Lastly, Hummel details the most common metric, the Euclidean distance, whereby the squared magnitude of the difference of squared vectors is summed, generating a score for matching purposes. It is of the interest of this article to elucidate potential additives resident within MPLs from a subset of the FLOPP-E (data reprinted with permission from [ 29 ]. Copyright 2021 American Chemical Society) and SLOPP-E (data reprinted with permission from [ 30 ]. Copyright 2020 American Chemical Society) database using a novel “Additive Analysis” Algorithm equipped to handle spectroscopic data from FTIR and Raman as well as their microspectroscopic modes. The algorithm is informed by additives from the Hummel [ 19 ] (available free to researchers) and Nava [ 31 ] (available free to researchers) databases. Moreover, we are proud to present an open-source digitized form of Hummel’s additive atlas (also made available in the SI). The discussion will illustrate the potential toxicological effects of the resulting additives elucidated by the algorithm. 2. METHODS 2.1 Code Structure The code will be split into 3 sections to make the script more legible and for the step-by-step guide to be easier to follow. The first section, “Data Preparation,” loads the numeric and string values from the databases of the user's choice while doing a user-defined calculation of peak wavenumbers if none are given previously [made up of subsections “Peak Finding of Unknown Spectra” and “Upload of DB (FTIR/Raman)”]. The second section, “Semi-quantitative Estimation,” calculates the relative scores from a simple metric based on the number of hits and their distance to the center of the window set by a user-specified tolerance. The third section, “Score Attribution,” assigns the name of the string header to the top ten scores identified in the database. Details on the data used in this paper, as well as the code, will be present in the SI. Data Preparation (Lines 9–52) format long swc = input("Is your unknown spectrum a collection of peaks? (Type: '1' for 'yes' or '2' for 'no'):"); Line 9: Formats output for longer output of scores for easier discrimination. Line 10: User-defined input that shifts execution towards peak-finding of spectra vs. redefining a pre-formatted array of peaks from the .mat file. if swc = = 2 Data = load("Plastic polymer controls\HDPE processed.csv"); %Change read according to data type (readmatrix for excel)! X_unk = Data(:,1); %Change name of variable accordingly (change back to dot index structure if .mat) Y_unk = 100 - Data(:,2); %Trans. to Abs. [pks, locs] = findpeaks(Y_unk,NPeaks = 20,SortStr="descend",MinPeakProminence = .15);%New Additions... Pk_XVals = zeros(length(locs),1); for i = 1:length(locs) Pk_XVals(i) = X_unk(locs(i)); end elseif swc = = 1 plchdr = load("Additive Algorithm Code + Misc\PVC_CP.mat"); %Uncomment if you have peak values already Pk_XVals = plchdr.PVC_CP; else disp("Input error therefore premature termination was executed.") return end Line 11: Multiple conditions for the two possible inputs and the null case. Lines 12–16: Firstly, the “unknown” samples are loaded in a .csv format. Next, the x-value array and y-value array are defined so that, later, peak-value locations can be used as an index to find the peak value and that the y-value array can be used as the principal variable in the “findpeaks” MATLAB function (Note: dot indexing must supplant loaded data in structure format). For the y-value array, the values are translated to “absorbance” inverting the y-value array (Note: This may not have to be executed if the code is already in absorbance or positive peaks. Also, you may have to institute rescaling or other normalization techniques if the spectra are not processed). For the “findpeaks” function, the y-value array is used as an input along with a set number of peaks (arbitrary choice of 20), a sort that takes the top 20 strongest peaks (most likely in terms of prominence), and a set tolerance for the minimum peak prominence avoiding the inclusion of noise and limiting the number of peaks generated (Note: The function outputs ALL maxima, therefore, adjust parameters accordingly). The outputs “pks” and “locs” refer to the y-value of the peak and the location of these values in the y-value array. Line 18–20: The for loop simply indexes the x-value array using the locations “locs: generated by the “findpeaks” function, resulting in the wave number location of the peaks versus their position in the array. Line 21–25: Alternative conditional statement where predetermined peaks (e.g., by visual confirmation or literary sources) are loaded and defined accordingly for the main body of the code. “.mat” does not always need dot indexing in the case of the file being a double , for example, but structure -formatted data requires it; therefore, the user may need to comment/bypass line 23. Line 24–27: Alternative input where the user errs, terminating the code prematurely. %% Upload of DB (FTIR/Raman) [D,S] = xlsread("Additive Algorithm Code + Misc\Wave Numbers of Common Plastic Additives.xlsx"); D(isnan(D)) = 0; %turns all NaN to zero entries D = max(D,0); %All negative entries are zero entries to avoid neg. minima being labeled as peaks NP = input("Is your database a collection of spectra, not peaks? (Type: '1' for 'yes' or '2' for 'No'):"): Line 30: Generates numerical [D] and string outputs [S] from a labeled Excel spreadsheet file in .xlsx format. Depending on the structure of the spreadsheet, some rows may have to be deleted to accommodate the code. D should be your spectra’s y-values or peaks, and [S] should be the associated name or chemical formula of the y-value column. Line 31: Renders all NaN entries as zero due to non-numerical data in the spreadsheet (headers/labels) Line 32: (For unprocessed spectra without min-max normalization) Renders negative y-values as 0. This goes for negative peaks. Line 33: User-specified input that shifts code into MATLAB’s in-built function for peak-finding. if NP = = 1 D_x = D(:,1); for i = 1:length(D(1,:))-1 [pks, locs] = findpeaks(D(:,i + 1),NPeaks = 20,SortStr="descend"); for j = 1:length(locs) E(j,i + 1) = D_x(locs(j)); end end E(:,1) = []; D = E; D = sort(D,"ascend"); S(:,1) = []; l = length(S(:,1)); S(2:l,:) = []; else disp("Input error therefore premature termination was executed.") return end Line 35: Condition for peak finding of the database. Lines 36–42: Similar to Lines 12–20. In this case, the database is a set of one x-value array describing the locations of intensities in spectroscopy by wavenumber (1/cm). The rest are y-value arrays belonging to entries in the database. The for loop is designed to run through these entries as a function to the number of columns minus 1 excluding counting of the x-value array. The “findpeaks” function is set up similarly to Lines 12–20, except a minimum prominence is not specified (not always necessary). The nested for loop simply indexes the locations using the x-value array “D_x” into a new “E,” preventing active overwriting of D being used as input over the cycle of the length of the dataset (j) and the entry number (i). Lines 43–48: Removes empty column array in lieu of x-value array. D is redefined, transforming the spectra into a set of peaks. D is sorted accordingly so that the main body of the code can step through the wavenumbers accordingly. S, the string matrix, is modified so that the x-value array is excised and everything is excluded except the header for the entries in the database. This allows for the correctly identified entry to be named at the tail-end of the code. Line 49–52: Termination of code is executed for alternative entries by the user. Semi-quantitative Estimation (Lines: 54–110) For the main body of the code, the line numbers will no longer be sequential. For ease, I will describe the outermost loops first. This is only in effect for lines 59 through 111. k = 1; h = 0; %number of hits on a given DB entry with respect to unknown peaks dif_norm = zeros(length(Pk_XVals),1); %In case there is no match within the calculations... ZeroPH = zeros(100,1); %creating a variable with arbitrary dimension for the conditional statement before assignment tol = input("What is your set tolerance (1/cm)?:"); Lines 54–57: Sets initial variables for the main body of code. “dif_norm” refers to the differences calculated by the code from the peaks found in the unknown sample to a particular entry in the database. “ZeroPH” is initialized allowing for peaks registered outside of the tolerance. Line 58: Tolerance set for the dif_norm calculation by the user. The paper used tolerances of 5, 10, and 15 1/cm. for i = 1:length(D(1,:)) %for length of columns (aka entries) for j = 1:length(D(:,i)) if D(j,i) = = 0 || j = = 20%choosing length end condition based on no zero entries Lines 59–61: The outermost for loop increments through the matrix of peak data sequentially. The next loop increments through the number of peaks in the ith column (database entry). The first conditional statement checks if the value of the database of peaks has reached 0 or if the first null entry is achieved while also looking for the 20th peak arbitrarily set for the unknown samples. This indicates that the score can be determined and tabulated accordingly. search_range = [D(j,i) - tol, D(j,i) + tol]; for k = 1:length(Pk_XVals) %scanning for peak matches in DB entry one unknown peak at a time if Pk_XVals(k) > = search_range(1) && Pk_XVals(k) < = search_range(2) if Pk_XVals(k) = = D(j,i) ZeroPH(k) = k; %Helps overwrite non-matches as perfect matches under this condition end dif = abs(D(j,i) - Pk_XVals(k)); dif_norm(k,1) = dif/tol; h = h + 1; %hit in DB entry elseif (Pk_XVals(k) - D(j,i)) > = 50%limits excessive iterating break else continue end end Line 94: Sets the search range or search window based on the value of the jth wavenumber for the ith entry in the databases plus or minus the tolerance set by the user. Lines 96–97: Sets outer for loop to terminate at the end of the length of peak values determined by previous sections of the code. The first conditional statement looks for peak values from the unknown within the search window of the database, including the values of the minimum and maximum values of the range. if Pk_XVals(k) = = D(j,i) ZeroPH(k) = k; %Helps overwrite non-matches as perfect match under this condition end Lines 98–100: The first conditional statement looks for peak values in the unknown sample that completely match the wave number in the database. If this is met, the index of the peak is set as a value instead of a difference of values between the “jth” row and “ith” column of D and the kth unknown peak value. Later on, this is redefined to differentiate between perfect and null entries. dif = abs(D(j,i) - Pk_XVals(k)); dif_norm(k,1) = dif/tol; h = h + 1; %hit in DB entry Lines 101–103: Difference calculated of D(j,i) and Pk_XVals(k) . The difference is normalized by the tolerance set by the user for easier score discrimination. The h index is incremented for easier score discrimination by counting the number of hits in the array of peak values with respect to the ith database entry. elseif (Pk_XVals(k) - D(j,i)) > = 50%limits excessive iterating break else continue end Lines 104–108: This alternative condition prevents the code from using the current search window for all peak values in the unknown spectra. Essentially, the code can move to the next search window, cutting down search time. The last three lines result in a zero entry for “dif_norm.” After these lines, the section loops through all peak values for the ith entry. Lines 62–77: Same lines from lines 95 through 108, which is reiterated due to the fact that the final calculation is not met for databases of raw spectra. It is unclear as to why this is occurring but it was a necessary measure for the code to run properly. h_tot = length(Pk_XVals); frac_scalar(i,1) = h/h_tot; dif_norm(dif_norm = = 0) = 15; %Setting max tolerance for zero Lines 78–80: Here, the scalar coefficient for the equation is calculated where the length of the array of peak values from the unknown normalizes the number of hits. The difference norm is redefined as “15,” where zero entries were placed for nonmatches indicating a failed search (since “15” is the maximum distance from the edge of the search window to the database peak under investigation). for m = 1:length(dif_norm) %entries equating to worst peak matches if ZeroPH(m) ~ = 0 dif_norm(m) = 0; %overwrites non-match placeholder as perfect match end end Lines 81–85: This for loop uses the index placement calculated on Line 98, giving the code the correct elements in the “dif_norm” to negate their assignment as the max tolerance. Perfect matches avoid an overwrite. SimScore(i,:) = (1/(mean(dif_norm))^2)*frac_scalar(i,1); %the smaller the dif the better h = 0; %Reinitialize components of score dif_norm = 0; ZeroPH = zeros(100,1); break Lines 87: Score determination takes place here, where the scores across the entire base are appended into a growing array. The score ( SS ) is calculated by taking the reciprocal of the average “dif_norm ( d-bar )” squared. This is scaled by the “frac_scalar ( f )” value, which penalizes the term by a hit ratio. The formulas are as follows: $$\:SS=\:\frac{1}{{{\stackrel{-}{d}}_{norm}}^{2}}f\left(i\right)$$ 1 Lines 88–90: The hit index, “dif_norm,” and “ZeroPH” are reinitialized so as to not alter further results. Score Attribution (Lines: 112–188) The final section of the code determines the top 10 scores from the database by indexing the database and string arrays using the similarity scores. The output is paired accordingly while controlling for duplicate and triplicate score values. SimScoreSort = sort(SimScore,'descend'); SimScoreSort(isnan(SimScoreSort)) = []; %Removed NaN values TopTenScores = SimScoreSort(1:10); i = 2; x = 1; %random value to keep the loop going y = 1; j = 2; %separate iterator for counting duplicates b = 0; %duplicate counter c = 0; %value of excess iterations from calculating duplicates e = 0; %placeholder for score skipped m = 1; %reorients for a duplicate at the top of the array Lines 112–114: Sorts scores from highest to lowest and selects for the top 10 scores while removing NaN values. Line 116–123: Various initial conditions used for various purposes. if i > 10 + c %Aids in terminating the loop break else Lines 125–127: Terminating condition. Left up for posterity as it was used to help terminate the loop when working with duplicates as opposed to duplicates and triplicates. if j − 1 = = 10 break else Lines 129–131: Avoids premature error throwing from code reaching the end of the “TopTenScores” array. if TopTenScores(j) = = TopTenScores(j-1) b = b + 1; j = j + 1; else j = j + 1; break end Lines 132–138: Checks for duplicate scores. If so, b is incremented such that multiple elements in the “OutputS” array can be assigned. J is incremented to run through the appropriate scores. if TopTenScores(i) = = TopTenScores(i-1) %This handles duplicate scores where one pulls 2 locations in the DB. ADJUST IF MORE THAN 2! TopTenPositions(i-1:i + b-1) = find(SimScore = = TopTenScores(i-1)); if NP = = 1 OutputS(i-1:i + b-1) = S(1,TopTenPositions(i-1:i + b-1)); %Raman DB has no title in header matrix i = i + 1 + b; b = 0; %reinitializing for next duplicate grouping, if any else OutputS(i-1:i + b-1) = S(2,TopTenPositions(i-1:i + b-1)); i = i + 1 + b; b = 0; %reinitializing for next duplicate grouping, if any end Lines 141–151: The first condition looks for duplicate/triplicate values. The first and second conditions in the nested if statement work similarly, aside from one difference. Firstly, “S” is indexed by grabbing the correct positions from i-1 to the “both” position. This is set in “OutputS” according to the number of duplicate/triplicate entries determined from b. “i” is incremented such that the code skips over to the next score or the next unequal score. “b” is reinitialized for counting duplicate presence. For “NP” equaling “2”, the string index starts on row 2. This is due to the nature of Nava’s headers in the Raman database. else TopTenPositions(i-1) = find(SimScore = = TopTenScores(i-1)); if NP = = 1 OutputS(i-1) = S(1,TopTenPositions(i-1)); i = i + 1; else OutputS(i-1) = S(2,TopTenPositions(i-1)); i = i + 1; end if length(TopTenPositions) = = length(TopTenScores) break end end Lines 154–167: Same assignment structure as Lines 141 through 151 for single score entries. if length(OutputS) ~ = 10%This means that the duplicate wasn't a part of the final value TopTenPositions(i-1) = 0; %Placeholder for array Z = find(SimScore = = TopTenScores(i-1)); %holds tenth or tenth, eleventh, etc. score TopTenPositions(i-1) = Z(1); %Final Step if NP = = 1 OutputS(i-1) = S(1,TopTenPositions(i-1)); else OutputS(i-1) = S(2,TopTenPositions(i-1)); end end Lines 172–182: Final calculations are made here, whereby the ultimate entry can be determined for the output. The last index is reset and used to find the appropriate score for any condition of NP. disp("Top 10 Hits in Chosen DB and their scores:") for i = 1:length(OutputS) Ranking = [i;cell2mat(OutputS(i));"Score:";TopTenScores(i)]; disp(Ranking) end Lines 184–188: Displays the properly labeled scores. 2. 2 Software and Hardware MATLAB version R2024a was used in the development of the code on an HP Pavilion x360 Convertible [14m-dw0xxx]. The process used was an Intel® Core(TM) i3-1005G1 CPU at 1.20GHz. The RAM has about 11.7 GB of usable space. The Windows 10 operating system is 64-bit. 2.3 Data Procurement High-density polyethylene (HDPE (25 µm – 800k MW, Magerial)), low-density polyethylene (LDPE (300 µm-Max Diameter, Goodfellow)), polypropylene (PP (25–85 µm, Polysciences)), polystyrene (PS (100 µm-5 mL, MilliporeSigma)), polyethylene terephthalate (PET (300 µm -Max Diameter, Goodfellow)), polymethylmethacrylate (PMMA (48 µm, Goodfellow)), polyamide-6 (PA6 (55 µm, Goodfellow)), polyvinyl chloride (PVC (Eastchem)), and cellulose acetate (CA (387 µm, Hawai’i Pacific University, Center for Marine Debris Research)), were deposited conservatively on gold slides. For PS, a film was formed on the gold slide used for the RaptIR µ-FTIR (ThermoFisher Scientific, Nicolet RaptIR + FTIR Microscope) system overnight for manual dispersion the next day. 100 samples were taken, and their means were calculated and used as a basis for the wave numbers in the FTIR controls. For the Raman controls, wavenumbers from various sources were selected [ 32 ][ 33 ][ 34 ][ 35 ][ 36 ][ 37 ][ 38 ][ 39 ]. The former part of the study uses principal wavenumber from two outside sources of copper phthalocyanine (CP) [FTIR] [ 40 ] [SDBS-No: -4418] and diketo-pyrrolo-pyrroles (DPP or 2PyPPB) [Raman] [ 41 ]. The latter part of the study involved randomly chosen samples from the FLOPP-E and SLOPP-E datasets. From lowest to highest entry, based on color, for each polymer class, a random number was drawn (i.e., A random number generator was set from 1 to 4 for four hypothetical entries: #3, #6, #9, and #11. 3 was drawn and thus only #9 was selected for). This was done to cut down on time and to eliminate bias in our choices. For the Raman portions of the study, further research was needed to determine the chemical identity of Nava’s results. Thus, a comprehensive online database of common plastic colorants was found. One limitation of Nava is that most colorants are detailed in the database versus the more comprehensive Hummel database. Hummel (digitized in this study) and Nava’s database is 327 and 788 entries long with the formatting slightly different: In Hummel’s case, the first row, or headers, are the labels for each entry and its set of peak wavenumbers and in Nava’s case, instead of wavenumbers there is the y-value array for each entry’s spectrum, and there is a 2-row header that outputs from the loading of the database. Both are column-wise. 3. RESULTS 3.1 Continued Study of “C2. Blue Fiber” and “Polyester 12. Red Fiber” Table 1 tabulates the results from the top 10 hits (via row) based on Hummel’s Atlas. The first, second, and third entries in each cell are the identified plastic additive at tolerances of 15, 10, and 5 1/cm, respectively. The first sample is a control of the additive in question, CP, where the 2nd through 10th entries are the same CP wavenumbers distributed amongst the wavenumbers of common polymers that MPLs are made up of CA, HDPE, LDPE, PP, PS, PET, PMMA, PA, PVC. The last entry is the FLOPP-E entry, C12. Blue Fiber. For the positive control, CP, the top 3 hits possess 3 separate entries within the Hummel database. A “metal-free” variety is outputted as the 6th hit for a tolerance of 10 1/cm. For the mixed samples, CA through PVC, the top 3 hits possess 3 separate entries within the Hummel database, partially excluding PMMA, whose 1st and 3rd hit, for a tolerance of 15 1/cm, is determined as zinc iron chloride (brown) and a sulfonamide vulcanization retarder. For the sample, “C2. Blue Fiber,” 3 colorant additives are identified with the CP entries in all other samples. The inorganic metal silicate is placed as the 4th hit at a tolerance of 15 1/cm, indanthrone is placed as the 5th hit at a tolerance of 10 1/cm, and the metal-free phthalocyanine is placed as the 9th hit at a tolerance of 10 1/cm. The latter hit appears to align with some pattern of “metal-free phthalocyanine,” especially for HDPE, PMMA, and PVC. Interestingly, the placement of the metal-free CP did not align with the CA control. However, due to the heterogeneity of the MPL and the general cellulosic nature of the particle, shifts in prediction strength may change. Looking more closely for a consistent pattern to determine the presence of the metal-free copper phthalocyanine (Fig. 1 ), a potential pattern has been brought forth through the value of the scores of the metal-free CP at their respective positions. On average, the controls produced a score of ~ .009, whereas the FLOPP-E sample was ~ .011, about a 22% increase. A Shapiro-Wilk test of normality was performed to determine the use of a parametric test that can statistically determine the credibility of FLOPP-E’s sample. A Shapiro-Wilk test of normality was performed that determined the score distribution for the metal-free phthalocyanine in the mixed and FLOPP-E entry subset was normally distributed at a W-test statistic of .9405 with a p-value of .6409. Thus, a one-sample t-test was administered using the CP control as a hypothetical mean in a t-test statistic of .4879 at a p-value of .6387 [CI(95%): ~.007,.011). Therefore, the distribution of scores is not statistically significant from the control. Looking at the top 3 results for the FLOPP-E sample, a majority of the hits are colorless plasticizers, lubricants, and whitening agents. Interestingly, the 2nd and 3rd highest colorless plasticizers on the list are phthalates. Table 2 tabulates the Raman-based results with respect to the Nava database. Similarly to Table 1 ’s design, Table 2 ’s control and mixed sample subset are associated with the suspected additive: DPP. The 2nd through 10th entries are CA through PVC. The last entry is the SLOPP-E sample “Polyester 12. Red Fiber.” Across the board (Fig. 2 ), various pyrroles appear to make up a majority of the hits for the control and mixed samples, excluding the SLOPP-E entry, with the chemical family being represented 60 to 90% of the time regardless of tolerance with PET, PVC, and PMMA appearing to be the most confounding (somewhat similar to the FTIR trial). Looking into this relationship further, a significance test was performed. Firstly, the Shapiro-Wilk test of normality was conducted sequentially until we found the distribution of the 10 1/cm tolerance to be significantly departed from normality (W: .8328, p: .04001). According to the Kruskal-Wallis test, the group of pyrrole IDs was statistically significantly different from each other (H: 9.3529, p: .00931), indicating in-consistent IDing across polymer types. For quinones (anthraquinones), IDs arose as a result in 6 out of 10 of the control and mixed samples, highlighting inconsistent IDing. This is fascinating in that the top 2 principal hits (tol: 5 1/cm) were registered as metal naphthoquinone (quinone), which share structural similarities. However, a vast majority of hits for the SLOPP-E entry were from the azo family. Looking further into comparing quinone IDs, regardless of tolerance across all samples in the trial, a significance test was performed to determine if there is a statistically significant difference between the scores in the control and mixed sample subset vs the quinone IDs in the SLOPP-E, entry. Again, the Shapiro-Wilk test of normality was conducted to determine the use of a non-parametric test on the distribution, where it was determined that the first group (control/mixed) and the second group (SLOPP-E) were both normally distributed with the caveat that the second group a sample size of 2. Despite normality, Mann-Whitney (Wilcoxon) was performed for unpaired groups of different sizes, where it was found that the groups were significantly different from the null of the means of both groups being equal (W: 0, p: .036364). For the IDs adjacent to red as a color, we wanted to see if there was any consistency in the proportion across all tolerances. Shapiro-Wilk test was performed to ascertain the use of a non-parametric test whereby no tolerance sub-group significantly departed from normality (W 5/10/15 : 9277/.868/.881 p 5/10/15 : .3983/.07751/.1111). Therefore, an ANOVA was performed, which determined no statistically significant changes in adjacent color ratio across all tolerances at an F-ratio test statistic of .01773 with a p-value of .982436. We also wanted to determine if there was general stability in the values of scores calculated regardless of mixing in addition to comparing scores from the sample polymers (Fig. 3 ). Figures 3 A and 3 B represent the FTIR trial distribution and the distribution without outliers, whereas Figs. 3 C and 3 D represent the Raman trial similarly. Due to the presence of extreme outliers, testing for any significant difference across score distributions with respect to polymer type was done via a Kruskal-Wallis test of medians. 11 groups were tested across all tolerances for both trials, resulting in an χ 2 -test statistic of 7.7504/3.9144 with a p-value of .6532/.9511, suggesting no statistically significant difference between score distributions across trials. Details on the score distribution for FTIR (Table S1 ) and Raman (Table S2) can be found in the supplementary information section of the publication. Note that a caveat of this analysis is that the “infinity” scores calculated were set to a 0 placeholder. This is generated when one or more wavenumbers in an unknown are a perfect match. 3.2 Extended Trial (Random Subset) Firstly, for the FTIR trial (Table 3 ), out of the 56 semi-randomly selected MPL samples from FLOPP-E, 49 samples generated IDs within the top ten scores calculated for the 788 entries tabulated from Hummel [ 19 ] and other sources, with some using a datapoint grab script from MATLAB [ 42 ][ 43 ][ 44 ][ 45 ][ 46 ][ 47 ][ 48 ][ 49 ][ 50 ](See Figure S1 for details). Of these 49, 28 were of the colorant type, whereas 21 were colorless additives. Interestingly, 77.6%, 10.2%, and 12.2% of the IDs arose from the 5, 10, and 15 1/cm tolerance wavenumber. The best matches in the 5 1/cm group were for 2 simultaneous entries, “PE35_BlackFragment” and “PE40_ClearFragment”. Their primary constituent was predicted to be benzenesulfonohydrazide, a colorless blowing agent at a score of .0051. For the 10 1/cm group, the entry “PP12_WhiteFragment” had a primary constituent predicted as titanium dioxide, a white colorant. For the 15 1/cm group, the entry, “PE10_BlueFragment”, had a primary constituent predicted as barium permanganate and barium sulfate, an inorganic blue colorant. Secondly, for the Raman trial (Table 4 ), 27 out of 40 semi-randomly selected ML samples from SLOPP-E generated plausible predictions in the top ten scores of the 327 entries from Nava [ 31 ]. Interestingly, 74.1%, 22.2%, and 3.7% of the matches generated were from the 5, 10, and 15 1/cm tolerance groups. For the 5 1/cm group, the entry, “Polyester7_PinkFiber”, had a primary constituent predicted as PR210, a red azo colorant at a score of .000776. For the 10 1/cm group, the entry, “Polypropylene10_BrownFiber”, had a primary constituent predicted as PBr25, a brown azo colorant at a score of .001435. For the 15 1/cm group, the entry, “CelluloseAcetate1_ClearFilm”, had a primary constituent predicted as muscovite, a typically colorless inorganic additive at a score of .001889. The scores, on average, are larger for the FTIR than the Raman trial. These top 6 (7 for barium salt)’s potential toxicity is left for the discussion . In Fig. 4 , the distribution of hits across the FTIR extended trial (Fig. 4 A) detects the presence of organic colorants, inorganic colorants, and blowing agents the majority of the time. In contrast, the Raman extended trial (Fig. 4 B) detects azo colorants, minerals, phthalocyanines, and triarylmethanes the majority of the time. 4. DISCUSSION Interestingly, for the FLOPP-E entry, the other two blue hits, indanthrone and metal-free phthalocyanine, are additives outside of the initial prediction made in our previous report, with the exception of the latter additive’s relatedness to CP. Theoretically, CP may lose its coordinate Cu 2+ in acidic conditions (especially in context to ocean acidification). If this prediction holds, this may be the first evidence of the dechelation phthalocyanine colorant in an environmental sample adding another diagnostic marker for environmentally-degraded MPLs. According to the analysis in Fig. 1 , the plausibility is there. For the colorless additives in “C2. Blue Fiber’s” predictions, due to their colorless behavior, it is unsafe to assume that these materials may not be present. Moreover, the concentration of such materials may be higher than the blue pigment resident within the particle, potentially due to different rates of leeching during environmental degradation. Interestingly, some of these predictions are phthalates similar to the metal-complex colorant, potentially giving more credence to the presence of phthalocyanine through obscured wavenumbers or porphyrin ring-opening events when pairing the metal-free variety argument. For the SLOPP-E trial, pyrrole predictions were stable across the pure sample control as well as the mixed controls. Fascinatingly, according to the algorithm, we were incorrect in our prediction of “Polyester 12. Red Fiber” possessing a pyrrole as its primary colorant. Instead, the azo and quinone family came to prominence. We wanted to see if this was connected to the few instances where quinones were identified in the controls possessed similar levels of scoring. Strikingly, the scoring in the SLOPP-E superseded scoring in the controls making it unlikely the entries are tied to pyrroles in some way. Moreover, azo-derivatives dominated identification across the entry, leading to the conclusion that a red azo of some variety may be the principal colorant as opposed to the quinone and metal quinone entries' principal identification. In our previous research, there was some evidence anthraquinones could be within the MPL matrix so its misidentification cannot be ruled out. We also studied if it was coincidental that a high proportion of predictions appeared to be the same for red or red-adjacent colors (orange and brown). According to statistical analysis, the prediction algorithm offers consistency in at least identifying the correct color in the sample. However, Nava’s database appears to be dominated by entries of this type: ~58.4%, which may bias the results. In the extended study, we saw a unique array of plausible additive predictions come up for both the FLOPP-E and SLOPP-E entries semi-randomly selected. Most of the predictions came to be within tighter tolerances than used in previous reports, 5 1/cm, potentially indicating more reliability in measurement. No statistical analyses were made, but a closer look into the structures of the top 6 additives with respect to human health was made. For potassium permanganate [ 51 ], a generic SDS profile details nausea, vomiting, and shortness of breath, among other symptoms, upon ingestion and general irritation for dermal contact. For barium sulfate [ 52 ], toxicological studies indicate little to no impact with ingestion due to its insoluble nature as a salt. It is excreted almost completely. For titanium dioxide [ 53 ][ 54 ], a wealth of literature has been conducted, which is encapsulated in an exhaustive review detailing potential carcinogenic effects (Type 2B Carcinogen), inflammation (cardiovascular/pulmonary), and potentially dermally, through catalyzing water, from water vapor in the atmosphere, into free radicals. This is mainly due to titania’s nature as a semiconductor. Moreover, there are also other theorized effects of long-term accumulation in the body. Muscovite [ 55 ] appears to exert its negative health impacts on the pulmonary system where fibrosis occurs from inhalation as mica. Due to MPL atmospheric deposition [ 56 ][ 57 ], this may be a plausible toxicological factor. Azo dyes, like PBr25 and PR210, appear [ 58 ] to have a remarkable exertion of toxicology in the GI system via azo-reductase-mediated toxicity from gut microbes. Some of the metabolites possessed a variety of hepatocarcinogenic or mutagenic nature. Strikingly, many organic dyes, like azo’s, are ubiquitous in synthetic food coloring. There are a few caveats to this algorithm. To start, it is only applied in the narrow context of plastic additives. Future research should include environmentally-relevant small molecules as well as the polymers themselves. In addition, we noticed that the string output of the algorithm truncated long outputs. It is in the user's best interest to save the string output after each run to reference the entry’s full name in the database of choice. Moreover, the algorithm has issues with samples containing numerous perfect matches alongside perfect matches for wavenumbers across the board. Scoring these matches in context to the rest of the sample distribution is an area of future research therefore, the user will have to notate the wavenumbers manually to determine its fitness for the prediction. Lastly, some samples for the FTIR portion of the extended study came up a majority of the time, which may lead to the false positive of the entries whose predictions are namely “Antimony nickel titanium oxide yellow,” “Sicotan Yellow,” “Viridian Green,” and other chromium antimonide-based inorganic pigments. Moreover, these inorganics possess lower amounts of critical wavenumbers; therefore, differences are weighed more in the averaging and should be accounted for in the future. A universal number of zero placeholders in databases up to the maximum number of peaks in a database may have to be implemented. 5. CONCLUSION The heterogeneity of MPLs is ubiquitously known in the literature. Efforts to detail the composition of these particles have yielded great results, mainly in the field of molecular analysis. Moreover, computational methods appear to be aiding in the elucidation of the MPLs matrices quickening the analysis of multiple particles. In order to remediate these particles, knowing the makeup is of vital importance; therefore, we presented a preliminary algorithm designed to handle numerical data detailing the peak information or spectra information of MPLs from user-specified sources. In our case, we used the existing database to further apply their work and study their acquired MPLs toward our goal of delineating MPL heterogeneity. With a general idea of the constituents of most MPLs, we can use techniques like enzymatic degradation to handle proper sustainability practices in remediation. Simply incinerating or capturing these particles isn’t enough to deal with the problem. We hope this proof-of-concept can pave the way towards relevant formulations that not only degrade plastic polymers but their additives which they are endowed upon, sustainably. Lastly, we hope this research is applied to adsorbents for MPLs. Declarations Ethics approval and consent to participate: Not applicable. Consent for publication: Not applicable. Funding: The authors utilize no sources of funding from grants. Conflicts of Interest: The authors declare no conflicts of interest. Data Available: Reference databases and controls are made available in the supplementary material section of the manuscript. The samples used from FLOPP-E and SLOPP-E can be found freely available in their parent manuscripts in our references. Code Availability: The algorithm presented in the paper is made available in .m and .pdf file format in the supplementary material section of the manuscript. Authors’ Contributions: S.A. worked on project management of the study. W.W. authored the manuscript’s direction, composed the code and the paper, and conducted the data and statistical analyses. References Rogers K. microplastics. Encycl Br 2024. Boucher J, Friot D. Primary Microplastics in the Oceans: a Global Evaluation of Sources. Gland, Switzerland: Gland, Switzerland: IUCN; 2017. 10.2305/IUCN.CH.2017.01.en . Ziani K, Ioniță-Mîndrican C-B, Mititelu M, Neacșu SM, Negrei C, Moroșan E, Drăgănescu D, Preda O-T, Microplastics. A Real Global Threat for Environment and Food Safety: A State of the Art Review. Nutrients. 2023;15. 10.3390/nu15030617 . Albergamo V, Wohlleben W, Plata DL. Photochemical weathering of polyurethane microplastics produced complex and dynamic mixtures of dissolved organic chemicals. Environ Sci Process Impacts. 2023;25:432–44. 10.1039/D2EM00415A . Chamas A, Moon H, Zheng J, Qiu Y, Tabassum T, Jang JH, Abu-Omar M, Scott SL, Suh S. Degradation Rates of Plastics in the Environment. ACS Sustain Chem Eng. 2020;8:3494–511. 10.1021/acssuschemeng.9b06635 . Decker C. Photodegradation of PVC. Degrad Stabilisation PVC. 1984;81–136. 10.1007/978-94-009-5618-6_3 . Li C, Quo J, Pang Y, Chen S, Wang J. Single underwater image restoration by blue-green channels dehazing and red channel correction, Shanghai, China,: 2016 IEEE International Conference on Acoustics, Speech and Signal, Processing. (ICASSP); 2016, pp. 1731–5. 10.1109/ICASSP.2016.7471973 Hu K, Zhou P, Yang Y, Hall T, Nie G, Yao Y, Xiaoguang D, Wang S. Degradation of Microplastics by a Thermal Fenton Reaction. ACS EST Eng. 2022;2:110–20. 10.1021/acsestengg.1c00323 . Manzoor S, Naqash N, Rashid G, Singh R. Plastic Material Degradation and Formation of Microplastic in the Environment: A Review. Mater Today Proc. 2022;56:3254–60. 10.1016/j.matpr.2021.09.379 . Danso D, Chow J, Streit WR. Plastics: Environmental and Biotechnological Perspectives on Microbial Degradation. Appl Environ Microbiol. 2019;85. 10.1128/AEM.01095-19 . Duan J, Bolan N, Li Y, Ding S, Atugoda T, Vithanage M, Sarkar B, Tsang DCW, Kirkham MB. Weathering of microplastics and interaction with other coexisting constituents in terrestrial and aquatic environments. Water Res. 2021;196:117011. 10.1016/j.watres.2021.117011 . Romera-Castillo C, Lucas A, Mallenco-Fornies R, Briones-Rizo M, Calvo E, Pelejero C. Abiotic plastic leaching contributes to ocean acidification. Sci Total Environ. 2023;854:158683. 10.1016/j.scitotenv.2022.158683 . Medyńska-Juraszek A, Jadhav B. Influence of Different Microplastic Forms on pH and Mobility of Cu2 + and Pb2 + in Soil. Molecules. 2022;27. 10.3390/molecules27051744 . Bergmann M, Gutow L, Klages M. Marine Anthropogenic Litter. Cham: Springer International Publishing; 2015. 10.1007/978-3-319-16510-3 . Shao H, Han Z, Krasteva N, Wang D. Identification of signaling cascade in the insulin signaling pathway in response to nanopolystyrene particles. Nanotoxicology. 2019;13:174–88. 10.1080/17435390.2018.1530395 . Besseling E, Foekema EM, Van Franeker JA, Leopold MF, Kühn S, Bravo Rebolledo EL, Heße E, Mielke L, IJzer J, Kamminga P, Koelmans AA. Microplastic in a macro filter feeder: Humpback whale Megaptera novaeangliae. Mar Pollut Bull. 2015;95:248–52. 10.1016/j.marpolbul.2015.04.007 . Leslie HA, van Velzen MJM, Brandsma SH, Vethaak AD, Garcia-Vallejo JJ, Lamoree MH. Discovery and quantification of plastic particle pollution in human blood. Environ Int. 2022;107199. 10.1016/J.ENVINT.2022.107199 . Marfella R, Prattichizzo F, Sardu C, Fulgenzi G, Graciotti L, Spadoni T, D’Onofrio N, Scisciola L, La Grotta R, Frigé C, Pellegrini V, Municinò M, Siniscalchi M, Spinetti F, Vigliotti G, Vecchione C, Carrizzo A, Accarino G, Squillante A, Spaziano G, Mirra D, Esposito R, Altieri S, Falco G, Fenti A, Galoppo S, Canzano S, Sasso FC, Matacchione G, Olivieri F, Ferraraccio F, Panarese I, Paolisso P, Barbato E, Lubritto C, Balestrieri ML, Mauro C, Caballero AE, Rajagopalan S, Ceriello A, D’Agostino B, Iovino P, Paolisso G. Microplastics and Nanoplastics in Atheromas and Cardiovascular Events. N Engl J Med. 2024;390:900–10. 10.1056/NEJMoa2309822 . Hummel DO. Atlas of Plastics Additives. Berlin Heidelberg: Springer; 2002. 10.1007/978-3-642-56211-2 . Massardo S, Verzola D, Alberti S, Caboni C, Santostefano M, Eugenio Verrina E, Angeletti A, Lugani F, Ghiggeri GM, Bruschi M, Candiano G, Rumeo N, Gentile M, Cravedi P, La Maestra S, Zaza G, Stallone G, Esposito P, Viazzi F, Mancianti N, La Porta E, Artini C. MicroRaman spectroscopy detects the presence of microplastics in human urine and kidney tissue. Environ Int. 2024;184:108444. 10.1016/j.envint.2024.108444 . Ullah S, Ahmad S, Guo X, Ullah S, Ullah S, Nabi G, Wanghe K. A review of the endocrine disrupting effects of micro and nano plastic and their associated chemicals in mammals. Front Endocrinol (Lausanne). 2022;13:1084236. 10.3389/fendo.2022.1084236 . Klein CE. In: Kufe DW, Pollock RE, Weichselbaum RR, editors. The Hypothalamic-Pituitary-Gonadal Axis. 6 ed. Holland-Frei Cancer Medicine; 2003. Campanale C, Massarelli C, Savino I, Locaputo V, Uricchio VF. A Detailed Review Study on Potential Effects of Microplastics and Additives of Concern on Human Health. Int J Environ Res Public Health. 2020;17. 10.3390/ijerph17041212 . Kettle A. Accelerated Solvent Extraction for Additives in Polymer Materials. 2016. https://assets.thermofisher.com/TFS-Assets/CMD/Reference-Materials/WP-71256-ASE-Additives-WP71256-EN.pdf . Accessed 25 Oct 2024. Rosso B, Bravo B, Gregoris E, Barbante C, Gambaro A, Corami F. Quantification and Chemical Characterization of Plastic Additives and Small Microplastics (< 100 µm) in Highway Road Dust. Toxics. 2023;11:936. 10.3390/toxics11110936 . Fries E, Dekiff JH, Willmeyer J, Nuelle M-T, Ebert M, Remy D. Identification of polymer types and additives in marine microplastic particles using pyrolysis-GC/MS and scanning electron microscopy. Environ Sci Process Impacts. 2013;15:1949. 10.1039/c3em00214d . Böke JS, Popp J, Krafft C. Optical photothermal infrared spectroscopy with simultaneously acquired Raman spectroscopy for two-dimensional microplastic identification. Sci Rep. 2022;12:18785. 10.1038/s41598-022-23318-2 . Kumar K, Giehl A, Patz C-D. Chemometric assisted Fourier Transform Infrared (FTIR) Spectroscopic analysis of fruit wine samples: Optimizing the initialization and convergence criteria in the non-negative factor analysis algorithm for developing a robust classification model. Spectrochim Acta Part Mol Biomol Spectrosc. 2019;209:22–31. 10.1016/j.saa.2018.10.024 . De Frond H, Rubinovitz R, Rochman CM. µATR-FTIR Spectral Libraries of Plastic Particles (FLOPP and FLOPP-e) for the Analysis of Microplastics. Anal Chem. 2021;93:15878–85. 10.1021/acs.analchem.1c02549 . Munno K, De Frond H, O’Donnell B, Rochman CM. Increasing the Accessibility for Characterizing Microplastics: Introducing New Application-Based and Spectral Libraries of Plastic Particles (SLoPP and SLoPP-E). Anal Chem. 2020;92:2443–51. 10.1021/acs.analchem.9b03626 . Nava V, Frezzotti ML, Leoni B. Raman Spectroscopy for the Analysis of Microplastics in Aquatic Systems. Appl Spectrosc. 2021;75:1341–57. 10.1177/00037028211043119 . Naim AF, Al, AlFannakh H, Arafat S, Ibrahim SS. Characterization of PVC/MWCNTs Nanocomposite: Solvent Blend. Sci Eng Compos Mater. 2019;27:55–64. 10.1515/secm-2020-0003 . Katsara K, Kenanakis G, Viskadourakis Z, Papadakis V. Polyethylene Migration from Food Packaging on Cheese Detected by Raman and Infrared (ATR/FT-IR) Spectroscopy. Mater (Basel). 2021;14:3872. 10.3390/ma14143872 . Furukawa T, Sato H, Kita Y, Matsukawa K, Yamaguchi H, Ochiai S, Siesler H, Ozaki Y. Molecular Structure, Crystallinity and Morphology of Polyethylene/Polypropylene Blends Studied by Raman Mapping, Scanning Electron Microscopy, Wide Angle X-Ray Diffraction, and Differential Scanning Calorimetry. Polym J. 2006;38:1127–36. 10.1295/polymj.PJ2006056 . Vieira MF, Bovolato AL, de Fonseca C, Izumi BG, Brolo CMS. A Direct Immunoassay Based on Surface-Enhanced Spectroscopy Using AuNP/PS-b-P2VP Nanocomposites. Sensors. 2023;23. 10.3390/s23104810 . Nielsen AS, Batchelder DN, Pyrz R. Estimation of crystallinity of isotactic polypropylene using Raman spectroscopy. Polym (Guildf). 2002;43:2671–6. 10.1016/S0032-3861(02)00053-8 . Chen J, Li J, Xu L, Hong W, Yang Y, Chen X. The Glass-Transition Temperature of Supported PMMA Thin Films with Hydrogen Bond/Plasmonic Interface. Polym (Basel). 2019;11. 10.3390/polym11040601 . Puchowicz D, Cieslak M. Raman Spectroscopy in the Analysis of Textile Structures. In: Pathak CS, Kumar S, editors. Recent Dev. At. Force Microsc. Raman Spectrosc. Mater. Charact. Rijeka: IntechOpen; 2021. 10.5772/intechopen.99731 . Sánchez-Márquez JA, Fuentes-Ramírez R, Cano-Rodríguez I, Gamiño-Arroyo Z, Rubio-Rosas E, Kenny JM, Rescignano N. Membrane Made of Cellulose Acetate with Polyacrylic Acid Reinforced with Carbon Nanotubes and Its Applicability for Chromium Removal. Int J Polym Sci. 2015;2015:320631. 10.1155/2015/320631 . National Institute of Advanced Industrial Science and Technology. SDBSWeb n.d. https://sdbs.db.aist.go.jp Luňák S, Vyňuchal J, Horáčková P, Frumarová B, Žák Z, Kučerík J, Salyk O. Structure and Raman spectra of pyridyl substituted diketo-pyrrolo-pyrrole isomers and polymorphs. J Mol Struct. 2010;983:39–47. 10.1016/j.molstruc.2010.08.026 . Johnson M, Ates M, Arslan Z, Farah I, Bogatu C. Assessment of Crystal Morphology on Uptake, Particle Dissolution, and Toxicity of Nanoscale Titanium Dioxide on Artemia Salina. J Nanotoxicology Nanomed. 2017;2:11–27. 10.4018/JNN.2017010102 . Al-Amin M, Dey S, Rashid T, Ashaduzzaman M, Shamsuddin S. Solar Assisted Photocatalytic Degradation of Reactive Azo Dyes in Presence of Anatase. Titanium Dioxide. 2016;2:14–21. Accessed 25 Oct 2024. Manzo M, Ahmed H, Nasrazadani S. Study on emission spectral lines of hematite and magnetite for purity’s differentiation. AIP Adv. 2020;10. 10.1063/1.5143413 . Mysak E. PY083 diarylide AADMC yellow 275–0570.TIF. CAMEO. 2020. https://cameo.mfa.org/wiki/File:PY083_diarylide_AADMC_yellow_275-0570.TIF . Accessed 25 Oct 2024. Quinacridone. GuideChem. 2023. https://www.guidechem.com/encyclopedia/quinacridone-dic7717.html#:~:text=Quinacridone%2C with the chemical formula,paints%2C inks%2C and dyes. Accessed 25 Oct 2024. PV 23 – dioxazine purple. CHSOS 2024. https://chsopensource.org/pv-23-dioxazine-purple/ . Accessed 25 Oct 2024. Doke JGRABIT. MATLAB. 2024. https://www.mathworks.com/matlabcentral/fileexchange/7173-grabit . Accessed 25 Oct 2024. PY 139 – isoindoline yellow. CHSOS 2024. https://chsopensource.org/py-139-isoindoline-yellow/ . Accessed 25 Oct 2024. Zappielo CD, Nanicuacua DM, Santos WNL, dos, Silva DLF da, Dall’Antônia LH, de Oliveira FM, Clausen DN, Tarley CRT. Solid Phase Extraction to On-Line Preconcentrate Trace Cadmium Using Chemically Modified Nano-Carbon Black with 3-Mercaptopropyltrimethoxysilane. J Braz Chem Soc. 2016. 10.5935/0103-5053.20160052 National Center for Biotechnology Information. PubChem Compound Summary for CID 24587, Barium permanganate 2024. https://pubchem.ncbi.nlm.nih.gov/compound/Barium-permanganate . Accessed 25 Oct 2024. Toxicological Profile for Barium and Barium Compounds. Agency Toxic Subst Dis Regist 2007. https://www.ncbi.nlm.nih.gov/books/NBK598777/ . Accessed 25 Oct 2024. Skocaj M, Filipic M, Petkovic J, Novak S. Titanium dioxide in our everyday life; is it safe? Radiol Oncol. 2011;45. 10.2478/v10019-011-0037-0 . Racovita AD. Titanium Dioxide: Structure, Impact, and Toxicity. Int J Environ Res Public Health. 2022;19:5681. 10.3390/ijerph19095681 . Holopainen M, Vallyathan V, Hedenborg M, Klockars M. Toxicity of Phlogopite and Muscovite. Vitro. Heal. Relat. Eff. Phyllosilicates. Berlin, Heidelberg: Springer Berlin Heidelberg; 1990. pp. 349–60. 10.1007/978-3-642-75124-0_30 . Cai L, Wang J, Peng J, Tan Z, Zhan Z, Tan X, Chen Q. Characteristic of microplastics in the atmospheric fallout from Dongguan city, China: preliminary research and first evidence. Environ Sci Pollut Res. 2017;24:24928–35. 10.1007/s11356-017-0116-x . Dris R, Gasperi J, Saad M, Mirande C, Tassin B. Synthetic fibers in atmospheric fallout: A source of microplastics in the environment? Mar Pollut Bull. 2016;104:290–3. 10.1016/j.marpolbul.2016.01.006 . Feng J. Toxicological significance of azo dye metabolism by human intestinal microbiota. Front Biosci. 2012;E4:568. 10.2741/400 . Tables Table 1 to 4 are available in the Supplementary Files section. Additional Declarations No competing interests reported. Supplementary Files Table14.docx SI.zip Cite Share Download PDF Status: Published Journal Publication published 05 Mar, 2025 Read the published version in Microplastics and Nanoplastics → Version 1 posted Editorial decision: Revision requested 11 Jan, 2025 Reviews received at journal 07 Jan, 2025 Reviewers agreed at journal 23 Dec, 2024 Reviews received at journal 22 Dec, 2024 Reviewers agreed at journal 29 Nov, 2024 Reviewers invited by journal 29 Nov, 2024 Editor assigned by journal 02 Nov, 2024 Submission checks completed at journal 30 Oct, 2024 First submitted to journal 25 Oct, 2024 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-5334015","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":373289455,"identity":"e10cf8c4-885e-4e3e-b21c-1ff325af034a","order_by":0,"name":"Wesley Allen Williams","email":"data:image/png;base64,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","orcid":"","institution":"North Carolina Agricultural and Technical State University","correspondingAuthor":true,"prefix":"","firstName":"Wesley","middleName":"Allen","lastName":"Williams","suffix":""},{"id":373289456,"identity":"a68e201d-50dd-4ddc-ac3d-29b94fa964fd","order_by":1,"name":"Shyam Aravamudhan","email":"","orcid":"","institution":"North Carolina Agricultural and Technical State University","correspondingAuthor":false,"prefix":"","firstName":"Shyam","middleName":"","lastName":"Aravamudhan","suffix":""}],"badges":[],"createdAt":"2024-10-25 17:38:10","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-5334015/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-5334015/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1186/s43591-025-00114-z","type":"published","date":"2025-03-05T15:57:44+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":68740483,"identity":"bca64e12-d513-4f1c-864c-b5b2806e4ecc","added_by":"auto","created_at":"2024-11-11 14:19:00","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":46504,"visible":true,"origin":"","legend":"\u003cp\u003eScore distribution of “metal-free phthalocyanine” across the positive controls, mixed controls, and “C2. Blue Fiber” from the FLOPP-E database.\u003c/p\u003e","description":"","filename":"OnlineFigure1.png","url":"https://assets-eu.researchsquare.com/files/rs-5334015/v1/c335bb64658b95fcda6b123e.png"},{"id":68740484,"identity":"5668f4b4-085d-4dc6-9bfb-6d5237cb10f6","added_by":"auto","created_at":"2024-11-11 14:19:00","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":67233,"visible":true,"origin":"","legend":"\u003cp\u003eDistribution of various proportions of pyrroles, quinones, scoring of quinones, and generally red colors. \u003cstrong\u003eTop Left:\u003c/strong\u003e Proportion of pyrrole IDs with respect to tolerance across all entries. \u003cstrong\u003eTop Right:\u003c/strong\u003e Proportion of quinone IDs with respect to tolerance across all entries. \u003cstrong\u003eBottom Left:\u003c/strong\u003e Scoring from quinones across all entries irrespective of their position as a hit and their tolerance level. \u003cstrong\u003eBottom Right:\u003c/strong\u003e Proportions of adjacent colors to red IDd in the distribution.\u003c/p\u003e","description":"","filename":"OnlineFigure2.png","url":"https://assets-eu.researchsquare.com/files/rs-5334015/v1/eeccf872a712ec22c37fe35a.png"},{"id":68740011,"identity":"d88bd26c-9e11-45a0-8e30-ec7b3bb84040","added_by":"auto","created_at":"2024-11-11 14:11:00","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":32102,"visible":true,"origin":"","legend":"\u003cp\u003eDistribution of scores across FTIR and Raman trials. \u003cstrong\u003eA.\u003c/strong\u003e Full distribution of FTIR-related scores for the FLOPP-E test. \u003cstrong\u003eB.\u003c/strong\u003e Outlier omitted for visual clarity. \u003cstrong\u003eC.\u003c/strong\u003e Full distribution of Raman-related scores for the SLOPP-E test.\u003cstrong\u003e D.\u003c/strong\u003e Outlier omitted for visual clarity. Note: 1-10, 11-20,…, and 100-110 represent 11 groups from the first trial.\u003c/p\u003e","description":"","filename":"OnlineFigure3.png","url":"https://assets-eu.researchsquare.com/files/rs-5334015/v1/5c540aa1e7532ce5d9c5e197.png"},{"id":68740010,"identity":"34ac518d-31bd-427f-8050-c47f56569c15","added_by":"auto","created_at":"2024-11-11 14:11:00","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":42773,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eA. \u003c/strong\u003eNumber of hits with respect to additive type in the second FTIR trial.\u003cstrong\u003e B. \u003c/strong\u003eNumber of hits with respect to colorant chemical family in the second Raman trial.\u003c/p\u003e","description":"","filename":"OnlineFigure4.png","url":"https://assets-eu.researchsquare.com/files/rs-5334015/v1/cb2193f48b20b6e4127c5e7c.png"},{"id":78192000,"identity":"865e2e78-f9a2-42aa-8ccd-933a19b8a56e","added_by":"auto","created_at":"2025-03-10 20:15:28","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":924920,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-5334015/v1/120e083d-4d2e-44cc-a1eb-5c88ded17c67.pdf"},{"id":68740008,"identity":"c6a7ebf2-b41b-4832-b873-0cd428117d54","added_by":"auto","created_at":"2024-11-11 14:11:00","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":57836,"visible":true,"origin":"","legend":"","description":"","filename":"Table14.docx","url":"https://assets-eu.researchsquare.com/files/rs-5334015/v1/e154f1c7c58decafee627a15.docx"},{"id":68740014,"identity":"fabdc36e-7a35-4e90-8b4a-c3cca9f88144","added_by":"auto","created_at":"2024-11-11 14:11:01","extension":"zip","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":12499138,"visible":true,"origin":"","legend":"","description":"","filename":"SI.zip","url":"https://assets-eu.researchsquare.com/files/rs-5334015/v1/c1a63a425d7cf298c7c7f0df.zip"}],"financialInterests":"No competing interests reported.","formattedTitle":"Semi-quantitative Computational Analysis of Plastic Additives in a FLOPP-E and SLOPP-E Database Subset","fulltext":[{"header":"1. INTRODUCTION","content":"\u003cp\u003eMPLs are microparticles derived from industry (primary) or derived from environmental breakdown (secondary). The former type of MPLs are typically utilized as \u0026ldquo;microbeads found in personal care products\u0026rdquo; [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e] for abrasive effects like shower gels [\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e] or are found in synthetic textiles (laundered nylon, polyester, etc.)[\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e][\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e], tires (via erosion from driving motor vehicles [\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]), and paint-based markings and resins (presumably acrylics from paint primers)[\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]. Some primary MPLs are made to adapt the material properties of other plastic products [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]. Secondary MPLs originate from environmental forces such as abrasion and photo-oxidation via chain scission events [\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e], like the Norrish reaction [\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e][\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e] which would vary as a function of the body of water\u0026rsquo;s depth [\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e]; thermal oxidation (primarily via chain unfolding [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e] which would vary environmentally based on the rate of the heating and surrounding medium [\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e]); microbial digestion [\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e]; weathering from water [\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e]; and pH (secondary effect: degradation in water is thought to acidify ocean environments [\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e] while other reports [\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e] specify alkalinization in a terrestrial environment due to additives within the MPLs). As an aside, pH could, in theory, affect degradation by way of preferentially selecting alcohols for oxonium ion formation (leaving group). Many of these particles are inducted into the hydrosphere via various anthropogenic activities (waste, storm runoff, and maritime activities [\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e]) where they cause potent toxicity to marine biodiversity [\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e] as well as higher-order consumers, like \u003cem\u003eMegaptera\u003c/em\u003e [\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e] or human beings [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e][\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eIt is obvious that the plastic polymer type may possess interactions in a living system, but are the plastic additives causing an effect in tandem? Indeed, some reports have indicated or theorized that the presence of organic molecules, like plasticizers, in the human body is a causative agent for toxicity. One such report indicated the presence of copper phthalocyanine and hematite, a common blue and red/black pigment in plastic [\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e], in human kidneys [\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e]! There are many forms of additives but phthalate esters, bisphenol A (BPA), and tributyltin have been selected to encapsulate the significance of this study.\u003c/p\u003e \u003cp\u003eChiefly, plasticizers, like phthalate esters, have been shown [\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e] to hypertrophy thyroid cells, increase thyroid activity, and disrupt the hypothalamic-pituitary-thyroid axis (responsible for response to stress and regulation of metabolism). For the pituitary system, in murine models, the dysregulation of the hypothalamic-pituitary-gonadal (HPG) axis is seen affecting the release of gonadotropin-releasing hormone, follicle-stimulating hormone, and luteinizing hormone (impacts spermatogenesis and ovulation) [\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e]. In terms of the reproductive system, in murine models, oxidative stress, altered sperm physiology, and anti-androgenic effects are seen [\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e] with phthalate exposure. Ovarian steroidogenesis effects and reduced luteinizing hormone are apparent [\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e], as well. Generally, phthalate esters have been shown to impact prenatal and postnatal development [\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e]. For BPA, alteration in T4 levels, oxidative damage, and antagonism have been shown [\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e] in murine models while also affecting the blood-testis-barrier (BTB), significantly disrupting the ability to select against certain small molecules entering the glands. Moreover, spermatozoan DNA damage is induced by BPA, as well while in the female reproductive system, progesterone and oestradiol are inhibited. In the pituitary region, hypothalamic inflammation can occur [\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e]. In general, in the past, BPA has been shown to increase childhood obesity and cardiovascular disease incidence [\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e]. Lastly, for Tributyltin, thyroid follicle reduction and steroidogenic enzyme activity are impacted, leading to impaired testosterone production and defective spermatozoa [\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e]. Many of these additives may potentially impact metabolic activity and the fecundity of mammalian species, including human beings. In general, tin exposure could cause cancer, gastrointestinal (GI) issues, and potential clastogenesis (breakage, loss, or rearrangement of chromosomes) [\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eCurrent efforts to characterize plastic additives within pristine and environmental MPLs in the literature range from spectroscopic and chromatographic techniques, sometimes used with rigorous extraction protocols. One such protocol is the Soxhlet extraction process (soxhelation). Soxhelation involves using the special Soxhlet apparatus, which helps condense evaporated solvents and siphon them back to the round bottom flask. This constant recycling minimizes the amount of solvent evaporated out of the apparatus. At the same time, the central vessel allows for the collection of the small molecule or substance under investigation after recondensation. The concentrated extract is further processed for spectroscopic and chromatographic techniques. A technical report [\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e] from ThermoFisher details the 6-hour soxhelation of PVC resulting in an essentially 100% recovery (actually\u0026thinsp;\u0026gt;\u0026thinsp;100%) of DOA (dioctyl adipate - plasticizer), TOP (trioctyl phosphate \u003cem\u003e\u0026ndash;\u003c/em\u003e flame retardant/plasticizer\u003cem\u003e)\u003c/em\u003e, DOP (dioctyl phthalate \u0026ndash; plasticizer), and TOTM (\u003cem\u003etris-\u003c/em\u003e(2-ethylhexyl)trimellitates \u0026ndash; PVC plasticizer) (confirmed via gas chromatography). Other reports have used characterization techniques along with simpler extraction of whole particles, such as oleophilic extraction, in the next report. One such report [\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e], utilizing \u0026micro;-FTIR characterization, determined the constituents of extracted MPLs and microparticles to be an array of polymers and additives. More specifically, they surmise that the incidence of particles primarily owing to plastic additives (UV stabilizers, lubricants, fillers, antioxidants, and colorants, to name a few) may correlate with MPL abundance in the environment. Lastly, for chromatographic techniques, a report [\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e] using pyrolysis gas chromatography/mass spectrometry determined the presence of various plasticizers (e.g., phthalates, benzaldehyde, and 2,4-di-\u003cem\u003etert\u003c/em\u003e-butylphenol), titania, barium, zinc, and sulfur.\u003c/p\u003e \u003cp\u003eIn addition to the additives presented, thus far, plastic polymers have a plethora of additive types that may cause some effect due to their molecular structure like plasticizers, flame retardants, antistatics, antioxidants, surfactants, vulcanizing agents, soaps, crosslinkers, peptizers, foaming agents, lubricants, fillers, curing agents, biocides, metal deactivators, antiozonants, anti-blocking agents, and colorants [\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e]. Potentially, computational methods may quicken the identification of these additives in MP matrices. Currently, a common metric of spectral matching is the hit-quality index (HQI). HQI is the dot product of two y-vectors (unknown/reference), squared, divided by the dot product of the squared vectors. This essentially measures covariance between the two samples, meaning perfectly parallel vectors will reach a perfect value of one, and a completely independent set of vectors will reach the lowest value of 0. This has been used [\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e] in a novel 2D-HQI technique using both Raman and FTIR HQIs of environmentally derived MPLs. Other reports [\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e] utilize aspects from linear algebra, like in one report where significant factors were determined by SVD, spanned by the minima and maxima of each variable in spectra, and the non-negative factor analysis is initiated which looks at the contribution of the factor and the sample. Iterations are run at a set limit to overwrite negative values. Essentially, the user will be able to see a relationship between the samples and variables. However, one caveat is that irrelevant sections must be trimmed as a pre-processing technique. According to Hummel [\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e], various other techniques apply a set tolerance to an individual band that arises as a match; thus, different bands are given no calculation. In other cases, the total number of bands in a spectrum that matches \u0026ldquo;pinged\u0026rdquo; is used as a metric to control for the match contribution or sometimes used with an arbitrary scaling factor. For sample and reference matching, \u0026ldquo;Lowry-Huppler\u0026rdquo; algorithms contain a four-fold whole-spectrum search which is the summation of: 1. The magnitude of the dot product of the spectrum and reference vectors, 2. The previous value, squared, 3 and 4. The same as 1 and 2 except that the tolerance is applied, presumably, scaling the vectors. All vectors must be normalized absorbance values. Lastly, Hummel details the most common metric, the Euclidean distance, whereby the squared magnitude of the difference of squared vectors is summed, generating a score for matching purposes.\u003c/p\u003e \u003cp\u003eIt is of the interest of this article to elucidate potential additives resident within MPLs from a subset of the FLOPP-E (data reprinted with permission from [\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e]. Copyright 2021 American Chemical Society) and SLOPP-E (data reprinted with permission from [\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e]. Copyright 2020 American Chemical Society) database using a novel \u0026ldquo;Additive Analysis\u0026rdquo; Algorithm equipped to handle spectroscopic data from FTIR and Raman as well as their microspectroscopic modes. The algorithm is informed by additives from the Hummel [\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e] (available free to researchers) and Nava [\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e] (available free to researchers) databases. Moreover, we are proud to present an open-source digitized form of Hummel\u0026rsquo;s additive atlas (also made available in the SI). The discussion will illustrate the potential toxicological effects of the resulting additives elucidated by the algorithm.\u003c/p\u003e"},{"header":"2. METHODS","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003e2.1 Code Structure\u003c/h2\u003e \u003cp\u003eThe code will be split into 3 sections to make the script more legible and for the step-by-step guide to be easier to follow. The first section, \u0026ldquo;Data Preparation,\u0026rdquo; loads the numeric and string values from the databases of the user's choice while doing a user-defined calculation of peak wavenumbers if none are given previously [made up of subsections \u0026ldquo;Peak Finding of Unknown Spectra\u0026rdquo; and \u0026ldquo;Upload of DB (FTIR/Raman)\u0026rdquo;]. The second section, \u0026ldquo;Semi-quantitative Estimation,\u0026rdquo; calculates the relative scores from a simple metric based on the number of hits and their distance to the center of the window set by a user-specified tolerance. The third section, \u0026ldquo;Score Attribution,\u0026rdquo; assigns the name of the string header to the top ten scores identified in the database. Details on the data used in this paper, as well as the code, will be present in the SI.\u003c/p\u003e \u003cp\u003eData Preparation (Lines 9\u0026ndash;52)\u003c/p\u003e \u003cp\u003eformat long\u003c/p\u003e \u003cp\u003eswc\u0026thinsp;=\u0026thinsp;input(\"Is your unknown spectrum a collection of peaks? (Type: '1' for 'yes' or '2' for 'no'):\");\u003c/p\u003e \u003cp\u003e \u003cul\u003e \u003cli\u003e \u003cp\u003eLine 9: Formats output for longer output of scores for easier discrimination.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eLine 10: User-defined input that shifts execution towards peak-finding of spectra vs. redefining a pre-formatted array of peaks from the .mat file.\u003c/p\u003e \u003c/li\u003e \u003c/ul\u003e \u003c/p\u003e \u003cp\u003eif swc\u0026thinsp;=\u0026thinsp;=\u0026thinsp;2\u003c/p\u003e \u003cp\u003eData\u0026thinsp;=\u0026thinsp;load(\"Plastic polymer controls\\HDPE processed.csv\"); %Change read according to data type (readmatrix for excel)!\u003c/p\u003e \u003cp\u003eX_unk\u0026thinsp;=\u0026thinsp;Data(:,1); %Change name of variable accordingly (change back to dot index structure if .mat)\u003c/p\u003e \u003cp\u003eY_unk\u0026thinsp;=\u0026thinsp;100 - Data(:,2); %Trans. to Abs.\u003c/p\u003e \u003cp\u003e[pks, locs]\u0026thinsp;=\u0026thinsp;findpeaks(Y_unk,NPeaks\u0026thinsp;=\u0026thinsp;20,SortStr=\"descend\",MinPeakProminence\u0026thinsp;=\u0026thinsp;.15);%New Additions...\u003c/p\u003e \u003cp\u003ePk_XVals\u0026thinsp;=\u0026thinsp;zeros(length(locs),1);\u003c/p\u003e \u003cp\u003efor i\u0026thinsp;=\u0026thinsp;1:length(locs)\u003c/p\u003e \u003cp\u003ePk_XVals(i)\u0026thinsp;=\u0026thinsp;X_unk(locs(i));\u003c/p\u003e \u003cp\u003eend\u003c/p\u003e \u003cp\u003eelseif swc\u0026thinsp;=\u0026thinsp;=\u0026thinsp;1\u003c/p\u003e \u003cp\u003eplchdr\u0026thinsp;=\u0026thinsp;load(\"Additive Algorithm Code\u0026thinsp;+\u0026thinsp;Misc\\PVC_CP.mat\"); %Uncomment if you have peak values already\u003c/p\u003e \u003cp\u003ePk_XVals\u0026thinsp;=\u0026thinsp;plchdr.PVC_CP;\u003c/p\u003e \u003cp\u003eelse\u003c/p\u003e \u003cp\u003edisp(\"Input error therefore premature termination was executed.\")\u003c/p\u003e \u003cp\u003ereturn\u003c/p\u003e \u003cp\u003eend\u003c/p\u003e \u003cp\u003e \u003cul\u003e \u003cli\u003e \u003cp\u003eLine 11: Multiple conditions for the two possible inputs and the null case.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eLines 12\u0026ndash;16: Firstly, the \u0026ldquo;unknown\u0026rdquo; samples are loaded in a .csv format. Next, the x-value array and y-value array are defined so that, later, peak-value locations can be used as an index to find the peak value and that the y-value array can be used as the principal variable in the \u0026ldquo;findpeaks\u0026rdquo; MATLAB function (Note: dot indexing must supplant loaded data in \u003cem\u003estructure\u003c/em\u003e format). For the y-value array, the values are translated to \u0026ldquo;absorbance\u0026rdquo; inverting the y-value array (Note: This may not have to be executed if the code is already in absorbance or positive peaks. Also, you may have to institute rescaling or other normalization techniques if the spectra are not processed). For the \u0026ldquo;findpeaks\u0026rdquo; function, the y-value array is used as an input along with a set number of peaks (arbitrary choice of 20), a sort that takes the top 20 strongest peaks (most likely in terms of prominence), and a set tolerance for the minimum peak prominence avoiding the inclusion of noise and limiting the number of peaks generated (Note: The function outputs ALL maxima, therefore, adjust parameters accordingly). The outputs \u0026ldquo;pks\u0026rdquo; and \u0026ldquo;locs\u0026rdquo; refer to the y-value of the peak and the location of these values in the y-value array.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eLine 18\u0026ndash;20: The for loop simply indexes the x-value array using the locations \u0026ldquo;locs: generated by the \u0026ldquo;findpeaks\u0026rdquo; function, resulting in the wave number location of the peaks versus their position in the array.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eLine 21\u0026ndash;25: Alternative conditional statement where predetermined peaks (e.g., by visual confirmation or literary sources) are loaded and defined accordingly for the main body of the code. \u0026ldquo;.mat\u0026rdquo; does not always need dot indexing in the case of the file being a \u003cem\u003edouble\u003c/em\u003e, for example, but \u003cem\u003estructure\u003c/em\u003e-formatted data requires it; therefore, the user may need to comment/bypass line 23.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eLine 24\u0026ndash;27: Alternative input where the user errs, terminating the code prematurely.\u003c/p\u003e \u003c/li\u003e \u003c/ul\u003e \u003c/p\u003e \u003cp\u003e%% Upload of DB (FTIR/Raman)\u003c/p\u003e \u003cp\u003e[D,S]\u0026thinsp;=\u0026thinsp;xlsread(\"Additive Algorithm Code\u0026thinsp;+\u0026thinsp;Misc\\Wave Numbers of Common Plastic Additives.xlsx\");\u003c/p\u003e \u003cp\u003eD(isnan(D))\u0026thinsp;=\u0026thinsp;0; %turns all NaN to zero entries\u003c/p\u003e \u003cp\u003eD\u0026thinsp;=\u0026thinsp;max(D,0); %All negative entries are zero entries to avoid neg. minima being labeled as peaks\u003c/p\u003e \u003cp\u003eNP\u0026thinsp;=\u0026thinsp;input(\"Is your database a collection of spectra, not peaks? (Type: '1' for 'yes' or '2' for 'No'):\"):\u003c/p\u003e \u003cp\u003e \u003cul\u003e \u003cli\u003e \u003cp\u003eLine 30: Generates numerical [D] and string outputs [S] from a labeled Excel spreadsheet file in .xlsx format. Depending on the structure of the spreadsheet, some rows may have to be deleted to accommodate the code. D should be your spectra\u0026rsquo;s y-values or peaks, and [S] should be the associated name or chemical formula of the y-value column.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eLine 31: Renders all NaN entries as zero due to non-numerical data in the spreadsheet (headers/labels)\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eLine 32: (For unprocessed spectra without min-max normalization) Renders negative y-values as 0. This goes for negative peaks.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eLine 33: User-specified input that shifts code into MATLAB\u0026rsquo;s in-built function for peak-finding.\u003c/p\u003e \u003c/li\u003e \u003c/ul\u003e \u003c/p\u003e \u003cp\u003eif NP\u0026thinsp;=\u0026thinsp;=\u0026thinsp;1\u003c/p\u003e \u003cp\u003eD_x\u0026thinsp;=\u0026thinsp;D(:,1);\u003c/p\u003e \u003cp\u003efor i\u0026thinsp;=\u0026thinsp;1:length(D(1,:))-1\u003c/p\u003e \u003cp\u003e[pks, locs]\u0026thinsp;=\u0026thinsp;findpeaks(D(:,i\u0026thinsp;+\u0026thinsp;1),NPeaks\u0026thinsp;=\u0026thinsp;20,SortStr=\"descend\");\u003c/p\u003e \u003cp\u003efor j\u0026thinsp;=\u0026thinsp;1:length(locs)\u003c/p\u003e \u003cp\u003eE(j,i\u0026thinsp;+\u0026thinsp;1)\u0026thinsp;=\u0026thinsp;D_x(locs(j));\u003c/p\u003e \u003cp\u003eend\u003c/p\u003e \u003cp\u003eend\u003c/p\u003e \u003cp\u003eE(:,1) = [];\u003c/p\u003e \u003cp\u003eD\u0026thinsp;=\u0026thinsp;E;\u003c/p\u003e \u003cp\u003eD\u0026thinsp;=\u0026thinsp;sort(D,\"ascend\");\u003c/p\u003e \u003cp\u003eS(:,1) = [];\u003c/p\u003e \u003cp\u003el\u0026thinsp;=\u0026thinsp;length(S(:,1));\u003c/p\u003e \u003cp\u003eS(2:l,:) = [];\u003c/p\u003e \u003cp\u003eelse\u003c/p\u003e \u003cp\u003edisp(\"Input error therefore premature termination was executed.\")\u003c/p\u003e \u003cp\u003ereturn\u003c/p\u003e \u003cp\u003eend\u003c/p\u003e \u003cp\u003e \u003cul\u003e \u003cli\u003e \u003cp\u003eLine 35: Condition for peak finding of the database.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eLines 36\u0026ndash;42: Similar to Lines 12\u0026ndash;20. In this case, the database is a set of one x-value array describing the locations of intensities in spectroscopy by wavenumber (1/cm). The rest are y-value arrays belonging to entries in the database. The for loop is designed to run through these entries as a function to the number of columns minus 1 excluding counting of the x-value array. The \u0026ldquo;findpeaks\u0026rdquo; function is set up similarly to Lines 12\u0026ndash;20, except a minimum prominence is not specified (not always necessary). The nested for loop simply indexes the locations using the x-value array \u0026ldquo;D_x\u0026rdquo; into a new \u0026ldquo;E,\u0026rdquo; preventing active overwriting of D being used as input over the cycle of the length of the dataset (j) and the entry number (i).\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eLines 43\u0026ndash;48: Removes empty column array in lieu of x-value array. D is redefined, transforming the spectra into a set of peaks. D is sorted accordingly so that the main body of the code can step through the wavenumbers accordingly. S, the string matrix, is modified so that the x-value array is excised and everything is excluded except the header for the entries in the database. This allows for the correctly identified entry to be named at the tail-end of the code.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eLine 49\u0026ndash;52: Termination of code is executed for alternative entries by the user.\u003c/p\u003e \u003c/li\u003e \u003c/ul\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eSemi-quantitative Estimation (Lines: 54\u0026ndash;110)\u003c/p\u003e \u003cp\u003eFor the main body of the code, the line numbers will no longer be sequential. For ease, I will describe the outermost loops first. \u003cb\u003eThis is only in effect for lines 59 through 111.\u003c/b\u003e\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003cp\u003ek\u0026thinsp;=\u0026thinsp;1;\u003c/p\u003e \u003cp\u003eh\u0026thinsp;=\u0026thinsp;0; %number of hits on a given DB entry with respect to unknown peaks\u003c/p\u003e \u003cp\u003edif_norm\u0026thinsp;=\u0026thinsp;zeros(length(Pk_XVals),1); %In case there is no match within the calculations...\u003c/p\u003e \u003cp\u003eZeroPH\u0026thinsp;=\u0026thinsp;zeros(100,1); %creating a variable with arbitrary dimension for the conditional statement before assignment\u003c/p\u003e \u003cp\u003etol\u0026thinsp;=\u0026thinsp;input(\"What is your set tolerance (1/cm)?:\");\u003c/p\u003e \u003cp\u003e \u003cul\u003e \u003cli\u003e \u003cp\u003eLines 54\u0026ndash;57: Sets initial variables for the main body of code. \u0026ldquo;dif_norm\u0026rdquo; refers to the differences calculated by the code from the peaks found in the unknown sample to a particular entry in the database. \u0026ldquo;ZeroPH\u0026rdquo; is initialized allowing for peaks registered outside of the tolerance.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eLine 58: Tolerance set for the dif_norm calculation by the user. The paper used tolerances of 5, 10, and 15 1/cm.\u003c/p\u003e \u003c/li\u003e \u003c/ul\u003e \u003c/p\u003e \u003cp\u003efor i\u0026thinsp;=\u0026thinsp;1:length(D(1,:)) %for length of columns (aka entries)\u003c/p\u003e \u003cp\u003efor j\u0026thinsp;=\u0026thinsp;1:length(D(:,i))\u003cdiv class=\"BlockQuote\"\u003e\u003cp\u003eif D(j,i)\u0026thinsp;=\u0026thinsp;=\u0026thinsp;0 || j\u0026thinsp;=\u0026thinsp;=\u0026thinsp;20%choosing length end condition based on no zero entries\u003c/p\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003eLines 59\u0026ndash;61: The outermost for loop increments through the matrix of peak data sequentially. The next loop increments through the number of peaks in the ith column (database entry). The first conditional statement checks if the value of the database of peaks has reached 0 or if the first null entry is achieved while also looking for the 20th peak arbitrarily set for the unknown samples. This indicates that the score can be determined and tabulated accordingly.\u003c/p\u003e \u003cp\u003esearch_range = [D(j,i) - tol, D(j,i)\u0026thinsp;+\u0026thinsp;tol];\u003c/p\u003e \u003cp\u003efor k\u0026thinsp;=\u0026thinsp;1:length(Pk_XVals) %scanning for peak matches in DB entry one unknown peak at a time\u003c/p\u003e \u003cp\u003eif Pk_XVals(k)\u0026thinsp;\u0026gt;\u0026thinsp;=\u0026thinsp;search_range(1) \u0026amp;\u0026amp; Pk_XVals(k)\u0026thinsp;\u0026lt;\u0026thinsp;=\u0026thinsp;search_range(2)\u003c/p\u003e \u003cp\u003eif Pk_XVals(k)\u0026thinsp;=\u0026thinsp;=\u0026thinsp;D(j,i)\u003cdiv class=\"BlockQuote\"\u003e\u003cp\u003eZeroPH(k)\u0026thinsp;=\u0026thinsp;k; %Helps overwrite non-matches as perfect matches under this condition\u003c/p\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003eend\u003c/p\u003e \u003cp\u003edif\u0026thinsp;=\u0026thinsp;abs(D(j,i) - Pk_XVals(k));\u003c/p\u003e \u003cp\u003edif_norm(k,1)\u0026thinsp;=\u0026thinsp;dif/tol;\u003c/p\u003e \u003cp\u003eh\u0026thinsp;=\u0026thinsp;h\u0026thinsp;+\u0026thinsp;1; %hit in DB entry\u003c/p\u003e \u003cp\u003eelseif (Pk_XVals(k) - D(j,i))\u0026thinsp;\u0026gt;\u0026thinsp;=\u0026thinsp;50%limits excessive iterating\u003c/p\u003e \u003cp\u003ebreak\u003c/p\u003e \u003cp\u003eelse\u003c/p\u003e \u003cp\u003econtinue\u003c/p\u003e \u003cp\u003eend\u003c/p\u003e \u003cp\u003eend\u003c/p\u003e \u003cp\u003e \u003cul\u003e \u003cli\u003e \u003cp\u003eLine 94: Sets the search range or search window based on the value of the jth wavenumber for the ith entry in the databases plus or minus the tolerance set by the user.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eLines 96\u0026ndash;97: Sets outer for loop to terminate at the end of the length of peak values determined by previous sections of the code. The first conditional statement looks for peak values from the unknown within the search window of the database, including the values of the minimum and maximum values of the range.\u003c/p\u003e \u003c/li\u003e \u003c/ul\u003e \u003c/p\u003e \u003cp\u003eif Pk_XVals(k)\u0026thinsp;=\u0026thinsp;=\u0026thinsp;D(j,i)\u003c/p\u003e \u003cp\u003eZeroPH(k)\u0026thinsp;=\u0026thinsp;k; %Helps overwrite non-matches as perfect match under this condition\u003c/p\u003e \u003cp\u003eend\u003c/p\u003e \u003cp\u003eLines 98\u0026ndash;100: The first conditional statement looks for peak values in the unknown sample that completely match the wave number in the database. If this is met, the index of the peak is set as a value instead of a difference of values between the \u0026ldquo;jth\u0026rdquo; row and \u0026ldquo;ith\u0026rdquo; column of D and the kth unknown peak value. Later on, this is redefined to differentiate between perfect and null entries.\u003c/p\u003e \u003cp\u003edif\u0026thinsp;=\u0026thinsp;abs(D(j,i) - Pk_XVals(k));\u003c/p\u003e \u003cp\u003edif_norm(k,1)\u0026thinsp;=\u0026thinsp;dif/tol;\u003c/p\u003e \u003cp\u003eh\u0026thinsp;=\u0026thinsp;h\u0026thinsp;+\u0026thinsp;1; %hit in DB entry\u003c/p\u003e \u003cp\u003eLines 101\u0026ndash;103: Difference calculated of \u003cem\u003eD(j,i)\u003c/em\u003e and \u003cem\u003ePk_XVals(k)\u003c/em\u003e. The difference is normalized by the tolerance set by the user for easier score discrimination. The h index is incremented for easier score discrimination by counting the number of hits in the array of peak values with respect to the ith database entry.\u003c/p\u003e \u003cp\u003eelseif (Pk_XVals(k) - D(j,i))\u0026thinsp;\u0026gt;\u0026thinsp;=\u0026thinsp;50%limits excessive iterating\u003c/p\u003e \u003cp\u003ebreak\u003c/p\u003e \u003cp\u003eelse\u003c/p\u003e \u003cp\u003econtinue\u003c/p\u003e \u003cp\u003eend\u003c/p\u003e \u003cp\u003e \u003cul\u003e \u003cli\u003e \u003cp\u003eLines 104\u0026ndash;108: This alternative condition prevents the code from using the current search window for all peak values in the unknown spectra. Essentially, the code can move to the next search window, cutting down search time. The last three lines result in a zero entry for \u0026ldquo;dif_norm.\u0026rdquo; After these lines, the section loops through all peak values for the ith entry.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eLines 62\u0026ndash;77: Same lines from lines 95 through 108, which is reiterated due to the fact that the final calculation is not met for databases of raw spectra. It is unclear as to why this is occurring but it was a necessary measure for the code to run properly.\u003c/p\u003e \u003c/li\u003e \u003c/ul\u003e \u003c/p\u003e \u003cp\u003eh_tot\u0026thinsp;=\u0026thinsp;length(Pk_XVals);\u003c/p\u003e \u003cp\u003efrac_scalar(i,1)\u0026thinsp;=\u0026thinsp;h/h_tot;\u003c/p\u003e \u003cp\u003edif_norm(dif_norm\u0026thinsp;=\u0026thinsp;=\u0026thinsp;0)\u0026thinsp;=\u0026thinsp;15; %Setting max tolerance for zero\u003c/p\u003e \u003cp\u003eLines 78\u0026ndash;80: Here, the scalar coefficient for the equation is calculated where the length of the array of peak values from the unknown normalizes the number of hits. The difference norm is redefined as \u0026ldquo;15,\u0026rdquo; where zero entries were placed for nonmatches indicating a failed search (since \u0026ldquo;15\u0026rdquo; is the maximum distance from the edge of the search window to the database peak under investigation).\u003c/p\u003e \u003cp\u003efor m\u0026thinsp;=\u0026thinsp;1:length(dif_norm) %entries equating to worst peak matches\u003c/p\u003e \u003cp\u003eif ZeroPH(m)\u0026thinsp;~\u0026thinsp;=\u0026thinsp;0\u003c/p\u003e \u003cp\u003edif_norm(m)\u0026thinsp;=\u0026thinsp;0; %overwrites non-match placeholder as perfect match\u003c/p\u003e \u003cp\u003eend\u003c/p\u003e \u003cp\u003eend\u003c/p\u003e \u003cp\u003eLines 81\u0026ndash;85: This for loop uses the index placement calculated on Line 98, giving the code the correct elements in the \u0026ldquo;dif_norm\u0026rdquo; to negate their assignment as the max tolerance. Perfect matches avoid an overwrite.\u003c/p\u003e \u003cp\u003eSimScore(i,:) = (1/(mean(dif_norm))^2)*frac_scalar(i,1); %the smaller the dif the better\u003c/p\u003e \u003cp\u003eh\u0026thinsp;=\u0026thinsp;0; %Reinitialize components of score\u003c/p\u003e \u003cp\u003edif_norm\u0026thinsp;=\u0026thinsp;0;\u003c/p\u003e \u003cp\u003eZeroPH\u0026thinsp;=\u0026thinsp;zeros(100,1);\u003c/p\u003e \u003cp\u003ebreak\u003c/p\u003e \u003cp\u003e \u003cul\u003e \u003cli\u003e \u003cp\u003eLines 87: Score determination takes place here, where the scores across the entire base are appended into a growing array. The score (\u003cem\u003eSS\u003c/em\u003e) is calculated by taking the reciprocal of the average \u0026ldquo;dif_norm (\u003cem\u003ed-bar\u003c/em\u003e)\u0026rdquo; squared. This is scaled by the \u0026ldquo;frac_scalar (\u003cem\u003ef\u003c/em\u003e)\u0026rdquo; value, which penalizes the term by a hit ratio. The formulas are as follows:\u003c/p\u003e \u003c/li\u003e \u003c/ul\u003e \u003cdiv id=\"Equ1\" class=\"Equation\"\u003e \u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ1\" name=\"EquationSource\"\u003e\n$$\\:SS=\\:\\frac{1}{{{\\stackrel{-}{d}}_{norm}}^{2}}f\\left(i\\right)$$\u003c/div\u003e \u003cdiv class=\"EquationNumber\"\u003e1\u003c/div\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cul\u003e \u003cli\u003e \u003cp\u003eLines 88\u0026ndash;90: The hit index, \u0026ldquo;dif_norm,\u0026rdquo; and \u0026ldquo;ZeroPH\u0026rdquo; are reinitialized so as to not alter further results.\u003c/p\u003e \u003c/li\u003e \u003c/ul\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eScore Attribution (Lines: 112\u0026ndash;188)\u003c/p\u003e \u003cp\u003eThe final section of the code determines the top 10 scores from the database by indexing the database and string arrays using the similarity scores. The output is paired accordingly while controlling for duplicate and triplicate score values.\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003cp\u003eSimScoreSort\u0026thinsp;=\u0026thinsp;sort(SimScore,'descend');\u003c/p\u003e \u003cp\u003eSimScoreSort(isnan(SimScoreSort)) = []; %Removed NaN values\u003c/p\u003e \u003cp\u003eTopTenScores\u0026thinsp;=\u0026thinsp;SimScoreSort(1:10);\u003c/p\u003e \u003cp\u003ei\u0026thinsp;=\u0026thinsp;2;\u003c/p\u003e \u003cp\u003ex\u0026thinsp;=\u0026thinsp;1; %random value to keep the loop going\u003c/p\u003e \u003cp\u003ey\u0026thinsp;=\u0026thinsp;1;\u003c/p\u003e \u003cp\u003ej\u0026thinsp;=\u0026thinsp;2; %separate iterator for counting duplicates\u003c/p\u003e \u003cp\u003eb\u0026thinsp;=\u0026thinsp;0; %duplicate counter\u003c/p\u003e \u003cp\u003ec\u0026thinsp;=\u0026thinsp;0; %value of excess iterations from calculating duplicates\u003c/p\u003e \u003cp\u003ee\u0026thinsp;=\u0026thinsp;0; %placeholder for score skipped\u003c/p\u003e \u003cp\u003em\u0026thinsp;=\u0026thinsp;1; %reorients for a duplicate at the top of the array\u003c/p\u003e \u003cp\u003e \u003cul\u003e \u003cli\u003e \u003cp\u003eLines 112\u0026ndash;114: Sorts scores from highest to lowest and selects for the top 10 scores while removing NaN values.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eLine 116\u0026ndash;123: Various initial conditions used for various purposes.\u003c/p\u003e \u003c/li\u003e \u003c/ul\u003e \u003c/p\u003e \u003cp\u003eif i\u0026thinsp;\u0026gt;\u0026thinsp;10\u0026thinsp;+\u0026thinsp;c %Aids in terminating the loop\u003c/p\u003e \u003cp\u003ebreak\u003c/p\u003e \u003cp\u003eelse\u003c/p\u003e \u003cp\u003eLines 125\u0026ndash;127: Terminating condition. Left up for posterity as it was used to help terminate the loop when working with duplicates as opposed to duplicates and triplicates.\u003c/p\u003e \u003cp\u003eif j \u0026minus;\u0026thinsp;1\u0026thinsp;=\u0026thinsp;=\u0026thinsp;10\u003c/p\u003e \u003cp\u003ebreak\u003c/p\u003e \u003cp\u003eelse\u003c/p\u003e \u003cp\u003eLines 129\u0026ndash;131: Avoids premature error throwing from code reaching the end of the \u0026ldquo;TopTenScores\u0026rdquo; array.\u003c/p\u003e \u003cp\u003eif TopTenScores(j)\u0026thinsp;=\u0026thinsp;=\u0026thinsp;TopTenScores(j-1)\u003c/p\u003e \u003cp\u003eb\u0026thinsp;=\u0026thinsp;b\u0026thinsp;+\u0026thinsp;1;\u003c/p\u003e \u003cp\u003ej\u0026thinsp;=\u0026thinsp;j\u0026thinsp;+\u0026thinsp;1;\u003c/p\u003e \u003cp\u003eelse\u003c/p\u003e \u003cp\u003ej\u0026thinsp;=\u0026thinsp;j\u0026thinsp;+\u0026thinsp;1;\u003c/p\u003e \u003cp\u003ebreak\u003c/p\u003e \u003cp\u003eend\u003c/p\u003e \u003cp\u003eLines 132\u0026ndash;138: Checks for duplicate scores. If so, b is incremented such that multiple elements in the \u0026ldquo;OutputS\u0026rdquo; array can be assigned. J is incremented to run through the appropriate scores.\u003c/p\u003e \u003cp\u003eif TopTenScores(i)\u0026thinsp;=\u0026thinsp;=\u0026thinsp;TopTenScores(i-1) %This handles duplicate scores where one pulls 2 locations in the DB. ADJUST IF MORE THAN 2!\u003c/p\u003e \u003cp\u003eTopTenPositions(i-1:i\u0026thinsp;+\u0026thinsp;b-1)\u0026thinsp;=\u0026thinsp;find(SimScore\u0026thinsp;=\u0026thinsp;=\u0026thinsp;TopTenScores(i-1));\u003c/p\u003e \u003cp\u003eif NP\u0026thinsp;=\u0026thinsp;=\u0026thinsp;1\u003cdiv class=\"BlockQuote\"\u003e\u003cp\u003eOutputS(i-1:i\u0026thinsp;+\u0026thinsp;b-1)\u0026thinsp;=\u0026thinsp;S(1,TopTenPositions(i-1:i\u0026thinsp;+\u0026thinsp;b-1)); %Raman DB has no title in header matrix\u003c/p\u003e\u003cp\u003ei\u0026thinsp;=\u0026thinsp;i\u0026thinsp;+\u0026thinsp;1\u0026thinsp;+\u0026thinsp;b;\u003c/p\u003e\u003cp\u003eb\u0026thinsp;=\u0026thinsp;0; %reinitializing for next duplicate grouping, if any\u003c/p\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003eelse\u003c/p\u003e \u003cp\u003eOutputS(i-1:i\u0026thinsp;+\u0026thinsp;b-1)\u0026thinsp;=\u0026thinsp;S(2,TopTenPositions(i-1:i\u0026thinsp;+\u0026thinsp;b-1));\u003c/p\u003e \u003cp\u003ei\u0026thinsp;=\u0026thinsp;i\u0026thinsp;+\u0026thinsp;1\u0026thinsp;+\u0026thinsp;b;\u003c/p\u003e \u003cp\u003eb\u0026thinsp;=\u0026thinsp;0; %reinitializing for next duplicate grouping, if any\u003c/p\u003e \u003cp\u003eend\u003c/p\u003e \u003cp\u003eLines 141\u0026ndash;151: The first condition looks for duplicate/triplicate values. The first and second conditions in the nested if statement work similarly, aside from one difference. Firstly, \u0026ldquo;S\u0026rdquo; is indexed by grabbing the correct positions from i-1 to the \u0026ldquo;both\u0026rdquo; position. This is set in \u0026ldquo;OutputS\u0026rdquo; according to the number of duplicate/triplicate entries determined from b. \u0026ldquo;i\u0026rdquo; is incremented such that the code skips over to the next score or the next unequal score. \u0026ldquo;b\u0026rdquo; is reinitialized for counting duplicate presence. For \u0026ldquo;NP\u0026rdquo; equaling \u0026ldquo;2\u0026rdquo;, the string index starts on row 2. This is due to the nature of Nava\u0026rsquo;s headers in the Raman database.\u003c/p\u003e \u003cp\u003eelse\u003c/p\u003e \u003cp\u003eTopTenPositions(i-1)\u0026thinsp;=\u0026thinsp;find(SimScore\u0026thinsp;=\u0026thinsp;=\u0026thinsp;TopTenScores(i-1));\u003c/p\u003e \u003cp\u003eif NP\u0026thinsp;=\u0026thinsp;=\u0026thinsp;1\u003cdiv class=\"BlockQuote\"\u003e\u003cp\u003eOutputS(i-1)\u0026thinsp;=\u0026thinsp;S(1,TopTenPositions(i-1));\u003c/p\u003e\u003cp\u003ei\u0026thinsp;=\u0026thinsp;i\u0026thinsp;+\u0026thinsp;1;\u003c/p\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003eelse\u003cdiv class=\"BlockQuote\"\u003e\u003cp\u003eOutputS(i-1)\u0026thinsp;=\u0026thinsp;S(2,TopTenPositions(i-1));\u003c/p\u003e\u003cp\u003ei\u0026thinsp;=\u0026thinsp;i\u0026thinsp;+\u0026thinsp;1;\u003c/p\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003eend\u003c/p\u003e \u003cp\u003eif length(TopTenPositions)\u0026thinsp;=\u0026thinsp;=\u0026thinsp;length(TopTenScores)\u003cdiv class=\"BlockQuote\"\u003e\u003cp\u003ebreak\u003c/p\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003eend\u003c/p\u003e \u003cp\u003eend\u003c/p\u003e \u003cp\u003eLines 154\u0026ndash;167: Same assignment structure as Lines 141 through 151 for single score entries.\u003c/p\u003e \u003cp\u003eif length(OutputS)\u0026thinsp;~\u0026thinsp;=\u0026thinsp;10%This means that the duplicate wasn't a part of the final value\u003c/p\u003e \u003cp\u003eTopTenPositions(i-1)\u0026thinsp;=\u0026thinsp;0; %Placeholder for array\u003c/p\u003e \u003cp\u003eZ\u0026thinsp;=\u0026thinsp;find(SimScore\u0026thinsp;=\u0026thinsp;=\u0026thinsp;TopTenScores(i-1)); %holds tenth or tenth, eleventh, etc. score\u003c/p\u003e \u003cp\u003eTopTenPositions(i-1)\u0026thinsp;=\u0026thinsp;Z(1); %Final Step\u003c/p\u003e \u003cp\u003eif NP\u0026thinsp;=\u0026thinsp;=\u0026thinsp;1\u003cdiv class=\"BlockQuote\"\u003e\u003cp\u003eOutputS(i-1)\u0026thinsp;=\u0026thinsp;S(1,TopTenPositions(i-1));\u003c/p\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003eelse\u003cdiv class=\"BlockQuote\"\u003e\u003cp\u003eOutputS(i-1)\u0026thinsp;=\u0026thinsp;S(2,TopTenPositions(i-1));\u003c/p\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003eend\u003c/p\u003e \u003cp\u003eend\u003c/p\u003e \u003cp\u003eLines 172\u0026ndash;182: Final calculations are made here, whereby the ultimate entry can be determined for the output. The last index is reset and used to find the appropriate score for any condition of NP.\u003c/p\u003e \u003cp\u003edisp(\"Top 10 Hits in Chosen DB and their scores:\")\u003c/p\u003e \u003cp\u003efor i\u0026thinsp;=\u0026thinsp;1:length(OutputS)\u003c/p\u003e \u003cp\u003eRanking = [i;cell2mat(OutputS(i));\"Score:\";TopTenScores(i)];\u003c/p\u003e \u003cp\u003edisp(Ranking)\u003c/p\u003e \u003cp\u003eend\u003c/p\u003e \u003cp\u003eLines 184\u0026ndash;188: Displays the properly labeled scores.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003e2. 2 Software and Hardware\u003c/h3\u003e\n\u003cp\u003eMATLAB version R2024a was used in the development of the code on an HP Pavilion x360 Convertible [14m-dw0xxx]. The process used was an Intel\u0026reg; Core(TM) i3-1005G1 CPU at 1.20GHz. The RAM has about 11.7 GB of usable space. The Windows 10 operating system is 64-bit.\u003c/p\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003e2.3 Data Procurement\u003c/h2\u003e \u003cp\u003eHigh-density polyethylene (HDPE (25 \u0026micro;m \u0026ndash; 800k MW, Magerial)), low-density polyethylene (LDPE (300 \u0026micro;m-Max Diameter, Goodfellow)), polypropylene (PP (25\u0026ndash;85 \u0026micro;m, Polysciences)), polystyrene (PS (100 \u0026micro;m-5 mL, MilliporeSigma)), polyethylene terephthalate (PET (300 \u0026micro;m -Max Diameter, Goodfellow)), polymethylmethacrylate (PMMA (48 \u0026micro;m, Goodfellow)), polyamide-6 (PA6 (55 \u0026micro;m, Goodfellow)), polyvinyl chloride (PVC (Eastchem)), and cellulose acetate (CA (387 \u0026micro;m, Hawai\u0026rsquo;i Pacific University, Center for Marine Debris Research)), were deposited conservatively on gold slides. For PS, a film was formed on the gold slide used for the RaptIR \u0026micro;-FTIR (ThermoFisher Scientific, Nicolet RaptIR\u0026thinsp;+\u0026thinsp;FTIR Microscope) system overnight for manual dispersion the next day. 100 samples were taken, and their means were calculated and used as a basis for the wave numbers in the FTIR controls. For the Raman controls, wavenumbers from various sources were selected [\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e][\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e][\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e][\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e][\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e][\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e][\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e][\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e]. The former part of the study uses principal wavenumber from two outside sources of copper phthalocyanine (CP) [FTIR] [\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e] [SDBS-No: -4418] and diketo-pyrrolo-pyrroles (DPP or 2PyPPB) [Raman] [\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e]. The latter part of the study involved randomly chosen samples from the FLOPP-E and SLOPP-E datasets. From lowest to highest entry, based on color, for each polymer class, a random number was drawn (i.e., A random number generator was set from 1 to 4 for four hypothetical entries: #3, #6, #9, and #11. 3 was drawn and thus only #9 was selected for). This was done to cut down on time and to eliminate bias in our choices. For the Raman portions of the study, further research was needed to determine the chemical identity of Nava\u0026rsquo;s results. Thus, a comprehensive online database of common plastic colorants was found. One limitation of Nava is that most colorants are detailed in the database versus the more comprehensive Hummel database. Hummel (digitized in this study) and Nava\u0026rsquo;s database is 327 and 788 entries long with the formatting slightly different: In Hummel\u0026rsquo;s case, the first row, or headers, are the labels for each entry and its set of peak wavenumbers and in Nava\u0026rsquo;s case, instead of wavenumbers there is the y-value array for each entry\u0026rsquo;s spectrum, and there is a 2-row header that outputs from the loading of the database. Both are column-wise.\u003c/p\u003e \u003c/div\u003e"},{"header":"3. RESULTS","content":"\u003cdiv id=\"Sec7\" class=\"Section2\"\u003e\n \u003ch2\u003e3.1 Continued Study of \u0026ldquo;C2. Blue Fiber\u0026rdquo; and \u0026ldquo;Polyester 12. Red Fiber\u0026rdquo;\u003c/h2\u003e\n \u003cp\u003eTable \u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003e tabulates the results from the top 10 hits (via row) based on Hummel\u0026rsquo;s Atlas. The first, second, and third entries in each cell are the identified plastic additive at tolerances of 15, 10, and 5 1/cm, respectively. The first sample is a control of the additive in question, CP, where the 2nd through 10th entries are the same CP wavenumbers distributed amongst the wavenumbers of common polymers that MPLs are made up of CA, HDPE, LDPE, PP, PS, PET, PMMA, PA, PVC. The last entry is the FLOPP-E entry, C12. Blue Fiber. For the positive control, CP, the top 3 hits possess 3 separate entries within the Hummel database. A \u0026ldquo;metal-free\u0026rdquo; variety is outputted as the 6th hit for a tolerance of 10 1/cm. For the mixed samples, CA through PVC, the top 3 hits possess 3 separate entries within the Hummel database, partially excluding PMMA, whose 1st and 3rd hit, for a tolerance of 15 1/cm, is determined as zinc iron chloride (brown) and a sulfonamide vulcanization retarder. For the sample, \u0026ldquo;C2. Blue Fiber,\u0026rdquo; 3 colorant additives are identified with the CP entries in all other samples. The inorganic metal silicate is placed as the 4th hit at a tolerance of 15 1/cm, indanthrone is placed as the 5th hit at a tolerance of 10 1/cm, and the metal-free phthalocyanine is placed as the 9th hit at a tolerance of 10 1/cm. The latter hit appears to align with some pattern of \u0026ldquo;metal-free phthalocyanine,\u0026rdquo; especially for HDPE, PMMA, and PVC. Interestingly, the placement of the metal-free CP did not align with the CA control. However, due to the heterogeneity of the MPL and the general cellulosic nature of the particle, shifts in prediction strength may change. Looking more closely for a consistent pattern to determine the presence of the metal-free copper phthalocyanine (Fig. \u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003e), a potential pattern has been brought forth through the value of the scores of the metal-free CP at their respective positions. On average, the controls produced a score of ~\u0026thinsp;.009, whereas the FLOPP-E sample was ~\u0026thinsp;.011, about a 22% increase. A Shapiro-Wilk test of normality was performed to determine the use of a parametric test that can statistically determine the credibility of FLOPP-E\u0026rsquo;s sample. A Shapiro-Wilk test of normality was performed that determined the score distribution for the metal-free phthalocyanine in the mixed and FLOPP-E entry subset was normally distributed at a W-test statistic of .9405 with a p-value of .6409. Thus, a one-sample t-test was administered using the CP control as a hypothetical mean in a t-test statistic of .4879 at a p-value of .6387 [CI(95%): ~.007,.011). Therefore, the distribution of scores is not statistically significant from the control.\u003c/p\u003e\n \u003cp\u003eLooking at the top 3 results for the FLOPP-E sample, a majority of the hits are colorless plasticizers, lubricants, and whitening agents. Interestingly, the 2nd and 3rd highest colorless plasticizers on the list are phthalates.\u003c/p\u003e\n \u003cp\u003eTable \u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003e tabulates the Raman-based results with respect to the Nava database. Similarly to Table \u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003e\u0026rsquo;s design, Table \u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003e\u0026rsquo;s control and mixed sample subset are associated with the suspected additive: DPP. The 2nd through 10th entries are CA through PVC. The last entry is the SLOPP-E sample \u0026ldquo;Polyester 12. Red Fiber.\u0026rdquo; Across the board (Fig. \u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003e), various pyrroles appear to make up a majority of the hits for the control and mixed samples, excluding the SLOPP-E entry, with the chemical family being represented 60 to 90% of the time regardless of tolerance with PET, PVC, and PMMA appearing to be the most confounding (somewhat similar to the FTIR trial). Looking into this relationship further, a significance test was performed. Firstly, the Shapiro-Wilk test of normality was conducted sequentially until we found the distribution of the 10 1/cm tolerance to be significantly departed from normality (W: .8328, p: .04001). According to the Kruskal-Wallis test, the group of pyrrole IDs was statistically significantly different from each other (H: 9.3529, p: .00931), indicating in-consistent IDing across polymer types. For quinones (anthraquinones), IDs arose as a result in 6 out of 10 of the control and mixed samples, highlighting\u003c/p\u003e\n \u003cp\u003einconsistent IDing. This is fascinating in that the top 2 principal hits (tol: 5 1/cm) were registered as metal naphthoquinone (quinone), which share structural similarities. However, a vast majority of hits for the SLOPP-E entry were from the azo family. Looking further into comparing quinone IDs, regardless of tolerance across all samples in the trial, a significance test was performed to determine if there is a statistically significant difference between the scores in the control and mixed sample subset vs the quinone IDs in the SLOPP-E, entry. Again, the Shapiro-Wilk test of normality was conducted to determine the use of a non-parametric test on the distribution, where it was determined that the first group (control/mixed) and the second group (SLOPP-E) were both normally distributed with the caveat that the second group a sample size of 2. Despite normality, Mann-Whitney (Wilcoxon) was performed for unpaired groups of different sizes, where it was found that the groups were significantly different from the null of the means of both groups being equal (W: 0, p: .036364). For the IDs adjacent to red as a color, we wanted to see if there was any consistency in the proportion across all tolerances. Shapiro-Wilk test was performed to ascertain the use of a non-parametric test whereby no tolerance sub-group significantly departed from normality (W\u003csub\u003e5/10/15\u003c/sub\u003e: 9277/.868/.881 p\u003csub\u003e5/10/15\u003c/sub\u003e: .3983/.07751/.1111). Therefore, an ANOVA was performed, which determined no statistically significant changes in adjacent color ratio across all tolerances at an F-ratio test statistic of .01773 with a p-value of .982436.\u003c/p\u003e\n \u003cp\u003eWe also wanted to determine if there was general stability in the values of scores calculated regardless of mixing in addition to comparing scores from the sample polymers (Fig. \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003e). Figures \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003eA and \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003eB represent the FTIR trial distribution and the distribution without outliers, whereas Figs. \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003eC and \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003eD represent the Raman trial similarly. Due to the presence of extreme outliers, testing for any significant difference across score distributions with respect to polymer type was done via a Kruskal-Wallis test of medians. 11 groups were tested across all tolerances for both trials, resulting in an \u0026chi;\u003csup\u003e2\u003c/sup\u003e-test statistic of 7.7504/3.9144 with a p-value of .6532/.9511, suggesting no statistically significant difference between score distributions across trials. Details on the score distribution for FTIR (Table \u003cspan class=\"InternalRef\"\u003eS1\u003c/span\u003e) and Raman (Table S2) can be found in the supplementary information section of the publication. Note that a caveat of this analysis is that the \u0026ldquo;infinity\u0026rdquo; scores calculated were set to a 0 placeholder. This is generated when one or more wavenumbers in an unknown are a perfect match.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec8\" class=\"Section2\"\u003e\n \u003ch2\u003e3.2 Extended Trial (Random Subset)\u003c/h2\u003e\n \u003cp\u003eFirstly, for the FTIR trial (Table \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003e), out of the 56 semi-randomly selected MPL samples from FLOPP-E, 49 samples generated IDs within the top ten scores calculated for the 788 entries tabulated from Hummel [\u003cspan class=\"CitationRef\"\u003e19\u003c/span\u003e] and other sources, with some using a datapoint grab script from MATLAB [\u003cspan class=\"CitationRef\"\u003e42\u003c/span\u003e][\u003cspan class=\"CitationRef\"\u003e43\u003c/span\u003e][\u003cspan class=\"CitationRef\"\u003e44\u003c/span\u003e][\u003cspan class=\"CitationRef\"\u003e45\u003c/span\u003e][\u003cspan class=\"CitationRef\"\u003e46\u003c/span\u003e][\u003cspan class=\"CitationRef\"\u003e47\u003c/span\u003e][\u003cspan class=\"CitationRef\"\u003e48\u003c/span\u003e][\u003cspan class=\"CitationRef\"\u003e49\u003c/span\u003e][\u003cspan class=\"CitationRef\"\u003e50\u003c/span\u003e](See Figure \u003cspan class=\"InternalRef\"\u003eS1\u003c/span\u003e for details). Of these 49, 28 were of the colorant type, whereas 21 were colorless additives. Interestingly, 77.6%, 10.2%, and 12.2% of the IDs arose from the 5, 10, and 15 1/cm tolerance wavenumber. The best matches in the 5 1/cm group were for 2 simultaneous entries, \u0026ldquo;PE35_BlackFragment\u0026rdquo; and \u0026ldquo;PE40_ClearFragment\u0026rdquo;. Their primary constituent was predicted to be benzenesulfonohydrazide, a colorless blowing agent at a score of .0051. For the 10 1/cm group, the entry \u0026ldquo;PP12_WhiteFragment\u0026rdquo; had a primary constituent predicted as titanium dioxide, a white colorant. For the 15 1/cm group, the entry, \u0026ldquo;PE10_BlueFragment\u0026rdquo;, had a primary constituent predicted as barium permanganate and barium sulfate, an inorganic blue colorant.\u003c/p\u003e\n \u003cp\u003eSecondly, for the Raman trial (Table \u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003e), 27 out of 40 semi-randomly selected ML samples from SLOPP-E generated plausible predictions in the top ten scores of the 327 entries from Nava [\u003cspan class=\"CitationRef\"\u003e31\u003c/span\u003e]. Interestingly, 74.1%, 22.2%, and 3.7% of the matches generated were from the 5, 10, and 15 1/cm tolerance groups. For the 5 1/cm group, the entry, \u0026ldquo;Polyester7_PinkFiber\u0026rdquo;, had a primary constituent predicted as PR210, a red azo colorant at a score of .000776. For the 10 1/cm group, the entry, \u0026ldquo;Polypropylene10_BrownFiber\u0026rdquo;, had a primary constituent predicted as PBr25, a brown azo colorant at a score of .001435. For the 15 1/cm group, the entry, \u0026ldquo;CelluloseAcetate1_ClearFilm\u0026rdquo;, had a primary constituent predicted as muscovite, a typically colorless inorganic additive at a score of .001889. The scores, on average, are larger for the FTIR than the Raman trial. These top 6 (7 for barium salt)\u0026rsquo;s potential toxicity is left for the \u003cem\u003ediscussion\u003c/em\u003e. In Fig. \u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003e, the distribution of hits across the FTIR extended trial (Fig. \u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003eA) detects the presence of organic colorants, inorganic colorants, and blowing agents the majority of the time. In contrast, the Raman extended trial (Fig. \u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003eB) detects azo colorants, minerals, phthalocyanines, and triarylmethanes the majority of the time.\u003c/p\u003e\n\u003c/div\u003e"},{"header":"4. DISCUSSION","content":"\u003cp\u003eInterestingly, for the FLOPP-E entry, the other two blue hits, indanthrone and metal-free phthalocyanine, are additives outside of the initial prediction made in our previous report, with the exception of the latter additive\u0026rsquo;s relatedness to CP. Theoretically, CP may lose its coordinate Cu\u003csup\u003e2+\u003c/sup\u003e in acidic conditions (especially in context to ocean acidification). If this prediction holds, this may be the first evidence of the dechelation phthalocyanine colorant in an environmental sample adding another diagnostic marker for environmentally-degraded MPLs. According to the analysis in Fig. \u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003e, the plausibility is there. For the colorless additives in \u0026ldquo;C2. Blue Fiber\u0026rsquo;s\u0026rdquo; predictions, due to their colorless behavior, it is unsafe to assume that these materials may not be present. Moreover, the concentration of such materials may be higher than the blue pigment resident within the particle, potentially due to different rates of leeching during environmental degradation. Interestingly, some of these predictions are phthalates similar to the metal-complex colorant, potentially giving more credence to the presence of phthalocyanine through obscured wavenumbers or porphyrin ring-opening events when pairing the metal-free variety argument.\u003c/p\u003e\n\u003cp\u003eFor the SLOPP-E trial, pyrrole predictions were stable across the pure sample control as well as the mixed controls. Fascinatingly, according to the algorithm, we were incorrect in our prediction of \u0026ldquo;Polyester 12. Red Fiber\u0026rdquo; possessing a pyrrole as its primary colorant. Instead, the azo and quinone family came to prominence. We wanted to see if this was connected to the few instances where quinones were identified in the controls possessed similar levels of scoring. Strikingly, the scoring in the SLOPP-E superseded scoring in the controls making it unlikely the entries are tied to pyrroles in some way. Moreover, azo-derivatives dominated identification across the entry, leading to the conclusion that a red azo of some variety may be the principal colorant as opposed to the quinone and metal quinone entries\u0026apos; principal identification. In our previous research, there was some evidence anthraquinones could be within the MPL matrix so its misidentification cannot be ruled out. We also studied if it was coincidental that a high proportion of predictions appeared to be the same for red or red-adjacent colors (orange and brown). According to statistical analysis, the prediction algorithm offers consistency in at least identifying the correct color in the sample. However, Nava\u0026rsquo;s database appears to be dominated by entries of this type: ~58.4%, which may bias the results.\u003c/p\u003e\n\u003cp\u003eIn the extended study, we saw a unique array of plausible additive predictions come up for both the FLOPP-E and SLOPP-E entries semi-randomly selected. Most of the predictions came to be within tighter tolerances than used in previous reports, 5 1/cm, potentially indicating more reliability in measurement. No statistical analyses were made, but a closer look into the structures of the top 6 additives with respect to human health was made. For potassium permanganate [\u003cspan class=\"CitationRef\"\u003e51\u003c/span\u003e], a generic SDS profile details nausea, vomiting, and shortness of breath, among other symptoms, upon ingestion and general irritation for dermal contact. For barium sulfate [\u003cspan class=\"CitationRef\"\u003e52\u003c/span\u003e], toxicological studies indicate little to no impact with ingestion due to its insoluble nature as a salt. It is excreted almost completely. For titanium dioxide [\u003cspan class=\"CitationRef\"\u003e53\u003c/span\u003e][\u003cspan class=\"CitationRef\"\u003e54\u003c/span\u003e], a wealth of literature has been conducted, which is encapsulated in an exhaustive review detailing potential carcinogenic effects (Type 2B Carcinogen), inflammation (cardiovascular/pulmonary), and potentially dermally, through catalyzing water, from water vapor in the atmosphere, into free radicals. This is mainly due to titania\u0026rsquo;s nature as a semiconductor. Moreover, there are also other theorized effects of long-term accumulation in the body. Muscovite [\u003cspan class=\"CitationRef\"\u003e55\u003c/span\u003e] appears to exert its negative health impacts on the pulmonary system where fibrosis occurs from inhalation as mica. Due to MPL atmospheric deposition [\u003cspan class=\"CitationRef\"\u003e56\u003c/span\u003e][\u003cspan class=\"CitationRef\"\u003e57\u003c/span\u003e], this may be a plausible toxicological factor. Azo dyes, like PBr25 and PR210, appear [\u003cspan class=\"CitationRef\"\u003e58\u003c/span\u003e] to have a remarkable exertion of toxicology in the GI system via azo-reductase-mediated toxicity from gut microbes. Some of the metabolites possessed a variety of hepatocarcinogenic or mutagenic nature. Strikingly, many organic dyes, like azo\u0026rsquo;s, are ubiquitous in synthetic food coloring.\u003c/p\u003e\n\u003cp\u003eThere are a few caveats to this algorithm. To start, it is only applied in the narrow context of plastic additives. Future research should include environmentally-relevant small molecules as well as the polymers themselves. In addition, we noticed that the string output of the algorithm truncated long outputs. It is in the user\u0026apos;s best interest to save the string output after each run to reference the entry\u0026rsquo;s full name in the database of choice. Moreover, the algorithm has issues with samples containing numerous perfect matches alongside perfect matches for wavenumbers across the board. Scoring these matches in context to the rest of the sample distribution is an area of future research therefore, the user will have to notate the wavenumbers manually to determine its fitness for the prediction. Lastly, some samples for the FTIR portion of the extended study came up a majority of the time, which may lead to the false positive of the entries whose predictions are namely \u0026ldquo;Antimony nickel titanium oxide yellow,\u0026rdquo; \u0026ldquo;Sicotan Yellow,\u0026rdquo; \u0026ldquo;Viridian Green,\u0026rdquo; and other chromium antimonide-based inorganic pigments. Moreover, these inorganics possess lower amounts of critical wavenumbers; therefore, differences are weighed more in the averaging and should be accounted for in the future. A universal number of zero placeholders in databases up to the maximum number of peaks in a database may have to be implemented.\u003c/p\u003e"},{"header":"5. CONCLUSION","content":"\u003cp\u003eThe heterogeneity of MPLs is ubiquitously known in the literature. Efforts to detail the composition of these particles have yielded great results, mainly in the field of molecular analysis. Moreover, computational methods appear to be aiding in the elucidation of the MPLs matrices quickening the analysis of multiple particles. In order to remediate these particles, knowing the makeup is of vital importance; therefore, we presented a preliminary algorithm designed to handle numerical data detailing the peak information or spectra information of MPLs from user-specified sources. In our case, we used the existing database to further apply their work and study their acquired MPLs toward our goal of delineating MPL heterogeneity. With a general idea of the constituents of most MPLs, we can use techniques like enzymatic degradation to handle proper sustainability practices in remediation. Simply incinerating or capturing these particles isn\u0026rsquo;t enough to deal with the problem. We hope this proof-of-concept can pave the way towards relevant formulations that not only degrade plastic polymers but their additives which they are endowed upon, sustainably. Lastly, we hope this research is applied to adsorbents for MPLs.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003eEthics approval and consent to participate: Not applicable.\u003c/p\u003e\n\u003cp\u003eConsent for publication: Not applicable.\u003c/p\u003e\n\u003cp\u003eFunding: The authors utilize no sources of funding from grants.\u003c/p\u003e\n\u003cp\u003eConflicts of Interest: The authors declare no conflicts of interest.\u003c/p\u003e\n\u003cp\u003eData Available: Reference databases and controls are made available in the supplementary material section of the manuscript. The samples used from FLOPP-E and SLOPP-E can be found freely available in their parent manuscripts in our references.\u003c/p\u003e\n\u003cp\u003eCode Availability: The algorithm presented in the\u0026nbsp;paper is made available in .m and .pdf file format in the supplementary material section of the manuscript.\u003c/p\u003e\n\u003cp\u003eAuthors\u0026rsquo; Contributions: S.A. worked on project management of the study. W.W. authored the manuscript\u0026rsquo;s direction, composed the code and the paper, and conducted the data and statistical analyses.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eRogers K. microplastics. Encycl Br 2024.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBoucher J, Friot D. Primary Microplastics in the Oceans: a Global Evaluation of Sources. Gland, Switzerland: Gland, Switzerland: IUCN; 2017. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.2305/IUCN.CH.2017.01.en\u003c/span\u003e\u003cspan address=\"10.2305/IUCN.CH.2017.01.en\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZiani K, Ioniță-M\u0026icirc;ndrican C-B, Mititelu M, Neacșu SM, Negrei C, Moroșan E, Drăgănescu D, Preda O-T, Microplastics. A Real Global Threat for Environment and Food Safety: A State of the Art Review. Nutrients. 2023;15. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3390/nu15030617\u003c/span\u003e\u003cspan address=\"10.3390/nu15030617\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAlbergamo V, Wohlleben W, Plata DL. Photochemical weathering of polyurethane microplastics produced complex and dynamic mixtures of dissolved organic chemicals. Environ Sci Process Impacts. 2023;25:432\u0026ndash;44. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1039/D2EM00415A\u003c/span\u003e\u003cspan address=\"10.1039/D2EM00415A\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChamas A, Moon H, Zheng J, Qiu Y, Tabassum T, Jang JH, Abu-Omar M, Scott SL, Suh S. Degradation Rates of Plastics in the Environment. ACS Sustain Chem Eng. 2020;8:3494\u0026ndash;511. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1021/acssuschemeng.9b06635\u003c/span\u003e\u003cspan address=\"10.1021/acssuschemeng.9b06635\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDecker C. Photodegradation of PVC. Degrad Stabilisation PVC. 1984;81\u0026ndash;136. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/978-94-009-5618-6_3\u003c/span\u003e\u003cspan address=\"10.1007/978-94-009-5618-6_3\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi C, Quo J, Pang Y, Chen S, Wang J. Single underwater image restoration by blue-green channels dehazing and red channel correction, Shanghai, China,: 2016 IEEE International Conference on Acoustics, Speech and Signal, Processing. (ICASSP); 2016, pp. 1731\u0026ndash;5. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1109/ICASSP.2016.7471973\u003c/span\u003e\u003cspan address=\"10.1109/ICASSP.2016.7471973\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHu K, Zhou P, Yang Y, Hall T, Nie G, Yao Y, Xiaoguang D, Wang S. Degradation of Microplastics by a Thermal Fenton Reaction. ACS EST Eng. 2022;2:110\u0026ndash;20. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1021/acsestengg.1c00323\u003c/span\u003e\u003cspan address=\"10.1021/acsestengg.1c00323\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eManzoor S, Naqash N, Rashid G, Singh R. Plastic Material Degradation and Formation of Microplastic in the Environment: A Review. Mater Today Proc. 2022;56:3254\u0026ndash;60. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.matpr.2021.09.379\u003c/span\u003e\u003cspan address=\"10.1016/j.matpr.2021.09.379\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDanso D, Chow J, Streit WR. Plastics: Environmental and Biotechnological Perspectives on Microbial Degradation. Appl Environ Microbiol. 2019;85. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1128/AEM.01095-19\u003c/span\u003e\u003cspan address=\"10.1128/AEM.01095-19\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDuan J, Bolan N, Li Y, Ding S, Atugoda T, Vithanage M, Sarkar B, Tsang DCW, Kirkham MB. Weathering of microplastics and interaction with other coexisting constituents in terrestrial and aquatic environments. Water Res. 2021;196:117011. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.watres.2021.117011\u003c/span\u003e\u003cspan address=\"10.1016/j.watres.2021.117011\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRomera-Castillo C, Lucas A, Mallenco-Fornies R, Briones-Rizo M, Calvo E, Pelejero C. Abiotic plastic leaching contributes to ocean acidification. Sci Total Environ. 2023;854:158683. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.scitotenv.2022.158683\u003c/span\u003e\u003cspan address=\"10.1016/j.scitotenv.2022.158683\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMedyńska-Juraszek A, Jadhav B. Influence of Different Microplastic Forms on pH and Mobility of Cu2\u0026thinsp;+\u0026thinsp;and Pb2\u0026thinsp;+\u0026thinsp;in Soil. Molecules. 2022;27. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3390/molecules27051744\u003c/span\u003e\u003cspan address=\"10.3390/molecules27051744\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBergmann M, Gutow L, Klages M. Marine Anthropogenic Litter. Cham: Springer International Publishing; 2015. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/978-3-319-16510-3\u003c/span\u003e\u003cspan address=\"10.1007/978-3-319-16510-3\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShao H, Han Z, Krasteva N, Wang D. Identification of signaling cascade in the insulin signaling pathway in response to nanopolystyrene particles. Nanotoxicology. 2019;13:174\u0026ndash;88. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1080/17435390.2018.1530395\u003c/span\u003e\u003cspan address=\"10.1080/17435390.2018.1530395\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBesseling E, Foekema EM, Van Franeker JA, Leopold MF, K\u0026uuml;hn S, Bravo Rebolledo EL, He\u0026szlig;e E, Mielke L, IJzer J, Kamminga P, Koelmans AA. Microplastic in a macro filter feeder: Humpback whale Megaptera novaeangliae. Mar Pollut Bull. 2015;95:248\u0026ndash;52. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.marpolbul.2015.04.007\u003c/span\u003e\u003cspan address=\"10.1016/j.marpolbul.2015.04.007\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLeslie HA, van Velzen MJM, Brandsma SH, Vethaak AD, Garcia-Vallejo JJ, Lamoree MH. Discovery and quantification of plastic particle pollution in human blood. Environ Int. 2022;107199. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/J.ENVINT.2022.107199\u003c/span\u003e\u003cspan address=\"10.1016/J.ENVINT.2022.107199\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMarfella R, Prattichizzo F, Sardu C, Fulgenzi G, Graciotti L, Spadoni T, D\u0026rsquo;Onofrio N, Scisciola L, La Grotta R, Frig\u0026eacute; C, Pellegrini V, Municin\u0026ograve; M, Siniscalchi M, Spinetti F, Vigliotti G, Vecchione C, Carrizzo A, Accarino G, Squillante A, Spaziano G, Mirra D, Esposito R, Altieri S, Falco G, Fenti A, Galoppo S, Canzano S, Sasso FC, Matacchione G, Olivieri F, Ferraraccio F, Panarese I, Paolisso P, Barbato E, Lubritto C, Balestrieri ML, Mauro C, Caballero AE, Rajagopalan S, Ceriello A, D\u0026rsquo;Agostino B, Iovino P, Paolisso G. Microplastics and Nanoplastics in Atheromas and Cardiovascular Events. N Engl J Med. 2024;390:900\u0026ndash;10. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1056/NEJMoa2309822\u003c/span\u003e\u003cspan address=\"10.1056/NEJMoa2309822\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHummel DO. Atlas of Plastics Additives. Berlin Heidelberg: Springer; 2002. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/978-3-642-56211-2\u003c/span\u003e\u003cspan address=\"10.1007/978-3-642-56211-2\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMassardo S, Verzola D, Alberti S, Caboni C, Santostefano M, Eugenio Verrina E, Angeletti A, Lugani F, Ghiggeri GM, Bruschi M, Candiano G, Rumeo N, Gentile M, Cravedi P, La Maestra S, Zaza G, Stallone G, Esposito P, Viazzi F, Mancianti N, La Porta E, Artini C. MicroRaman spectroscopy detects the presence of microplastics in human urine and kidney tissue. Environ Int. 2024;184:108444. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.envint.2024.108444\u003c/span\u003e\u003cspan address=\"10.1016/j.envint.2024.108444\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eUllah S, Ahmad S, Guo X, Ullah S, Ullah S, Nabi G, Wanghe K. A review of the endocrine disrupting effects of micro and nano plastic and their associated chemicals in mammals. Front Endocrinol (Lausanne). 2022;13:1084236. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3389/fendo.2022.1084236\u003c/span\u003e\u003cspan address=\"10.3389/fendo.2022.1084236\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKlein CE. In: Kufe DW, Pollock RE, Weichselbaum RR, editors. The Hypothalamic-Pituitary-Gonadal Axis. 6 ed. Holland-Frei Cancer Medicine; 2003.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCampanale C, Massarelli C, Savino I, Locaputo V, Uricchio VF. A Detailed Review Study on Potential Effects of Microplastics and Additives of Concern on Human Health. Int J Environ Res Public Health. 2020;17. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3390/ijerph17041212\u003c/span\u003e\u003cspan address=\"10.3390/ijerph17041212\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKettle A. Accelerated Solvent Extraction for Additives in Polymer Materials. 2016. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://assets.thermofisher.com/TFS-Assets/CMD/Reference-Materials/WP-71256-ASE-Additives-WP71256-EN.pdf\u003c/span\u003e\u003cspan address=\"https://assets.thermofisher.com/TFS-Assets/CMD/Reference-Materials/WP-71256-ASE-Additives-WP71256-EN.pdf\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Accessed 25 Oct 2024.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRosso B, Bravo B, Gregoris E, Barbante C, Gambaro A, Corami F. Quantification and Chemical Characterization of Plastic Additives and Small Microplastics (\u0026lt;\u0026thinsp;100 \u0026micro;m) in Highway Road Dust. Toxics. 2023;11:936. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3390/toxics11110936\u003c/span\u003e\u003cspan address=\"10.3390/toxics11110936\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFries E, Dekiff JH, Willmeyer J, Nuelle M-T, Ebert M, Remy D. Identification of polymer types and additives in marine microplastic particles using pyrolysis-GC/MS and scanning electron microscopy. Environ Sci Process Impacts. 2013;15:1949. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1039/c3em00214d\u003c/span\u003e\u003cspan address=\"10.1039/c3em00214d\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eB\u0026ouml;ke JS, Popp J, Krafft C. Optical photothermal infrared spectroscopy with simultaneously acquired Raman spectroscopy for two-dimensional microplastic identification. Sci Rep. 2022;12:18785. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41598-022-23318-2\u003c/span\u003e\u003cspan address=\"10.1038/s41598-022-23318-2\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKumar K, Giehl A, Patz C-D. Chemometric assisted Fourier Transform Infrared (FTIR) Spectroscopic analysis of fruit wine samples: Optimizing the initialization and convergence criteria in the non-negative factor analysis algorithm for developing a robust classification model. Spectrochim Acta Part Mol Biomol Spectrosc. 2019;209:22\u0026ndash;31. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.saa.2018.10.024\u003c/span\u003e\u003cspan address=\"10.1016/j.saa.2018.10.024\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDe Frond H, Rubinovitz R, Rochman CM. \u0026micro;ATR-FTIR Spectral Libraries of Plastic Particles (FLOPP and FLOPP-e) for the Analysis of Microplastics. Anal Chem. 2021;93:15878\u0026ndash;85. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1021/acs.analchem.1c02549\u003c/span\u003e\u003cspan address=\"10.1021/acs.analchem.1c02549\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMunno K, De Frond H, O\u0026rsquo;Donnell B, Rochman CM. Increasing the Accessibility for Characterizing Microplastics: Introducing New Application-Based and Spectral Libraries of Plastic Particles (SLoPP and SLoPP-E). Anal Chem. 2020;92:2443\u0026ndash;51. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1021/acs.analchem.9b03626\u003c/span\u003e\u003cspan address=\"10.1021/acs.analchem.9b03626\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNava V, Frezzotti ML, Leoni B. Raman Spectroscopy for the Analysis of Microplastics in Aquatic Systems. Appl Spectrosc. 2021;75:1341\u0026ndash;57. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1177/00037028211043119\u003c/span\u003e\u003cspan address=\"10.1177/00037028211043119\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNaim AF, Al, AlFannakh H, Arafat S, Ibrahim SS. Characterization of PVC/MWCNTs Nanocomposite: Solvent Blend. Sci Eng Compos Mater. 2019;27:55\u0026ndash;64. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1515/secm-2020-0003\u003c/span\u003e\u003cspan address=\"10.1515/secm-2020-0003\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKatsara K, Kenanakis G, Viskadourakis Z, Papadakis V. Polyethylene Migration from Food Packaging on Cheese Detected by Raman and Infrared (ATR/FT-IR) Spectroscopy. Mater (Basel). 2021;14:3872. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3390/ma14143872\u003c/span\u003e\u003cspan address=\"10.3390/ma14143872\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFurukawa T, Sato H, Kita Y, Matsukawa K, Yamaguchi H, Ochiai S, Siesler H, Ozaki Y. Molecular Structure, Crystallinity and Morphology of Polyethylene/Polypropylene Blends Studied by Raman Mapping, Scanning Electron Microscopy, Wide Angle X-Ray Diffraction, and Differential Scanning Calorimetry. Polym J. 2006;38:1127\u0026ndash;36. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1295/polymj.PJ2006056\u003c/span\u003e\u003cspan address=\"10.1295/polymj.PJ2006056\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVieira MF, Bovolato AL, de Fonseca C, Izumi BG, Brolo CMS. A Direct Immunoassay Based on Surface-Enhanced Spectroscopy Using AuNP/PS-b-P2VP Nanocomposites. Sensors. 2023;23. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3390/s23104810\u003c/span\u003e\u003cspan address=\"10.3390/s23104810\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNielsen AS, Batchelder DN, Pyrz R. Estimation of crystallinity of isotactic polypropylene using Raman spectroscopy. Polym (Guildf). 2002;43:2671\u0026ndash;6. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/S0032-3861(02)00053-8\u003c/span\u003e\u003cspan address=\"10.1016/S0032-3861(02)00053-8\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen J, Li J, Xu L, Hong W, Yang Y, Chen X. The Glass-Transition Temperature of Supported PMMA Thin Films with Hydrogen Bond/Plasmonic Interface. Polym (Basel). 2019;11. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3390/polym11040601\u003c/span\u003e\u003cspan address=\"10.3390/polym11040601\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePuchowicz D, Cieslak M. Raman Spectroscopy in the Analysis of Textile Structures. In: Pathak CS, Kumar S, editors. Recent Dev. At. Force Microsc. Raman Spectrosc. Mater. Charact. Rijeka: IntechOpen; 2021. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.5772/intechopen.99731\u003c/span\u003e\u003cspan address=\"10.5772/intechopen.99731\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eS\u0026aacute;nchez-M\u0026aacute;rquez JA, Fuentes-Ram\u0026iacute;rez R, Cano-Rodr\u0026iacute;guez I, Gami\u0026ntilde;o-Arroyo Z, Rubio-Rosas E, Kenny JM, Rescignano N. Membrane Made of Cellulose Acetate with Polyacrylic Acid Reinforced with Carbon Nanotubes and Its Applicability for Chromium Removal. Int J Polym Sci. 2015;2015:320631. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1155/2015/320631\u003c/span\u003e\u003cspan address=\"10.1155/2015/320631\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNational Institute of Advanced Industrial Science and Technology. SDBSWeb n.d. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://sdbs.db.aist.go.jp\u003c/span\u003e\u003cspan address=\"https://sdbs.db.aist.go.jp\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLuň\u0026aacute;k S, Vyňuchal J, Hor\u0026aacute;čkov\u0026aacute; P, Frumarov\u0026aacute; B, Ž\u0026aacute;k Z, Kučer\u0026iacute;k J, Salyk O. Structure and Raman spectra of pyridyl substituted diketo-pyrrolo-pyrrole isomers and polymorphs. J Mol Struct. 2010;983:39\u0026ndash;47. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.molstruc.2010.08.026\u003c/span\u003e\u003cspan address=\"10.1016/j.molstruc.2010.08.026\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJohnson M, Ates M, Arslan Z, Farah I, Bogatu C. Assessment of Crystal Morphology on Uptake, Particle Dissolution, and Toxicity of Nanoscale Titanium Dioxide on Artemia Salina. J Nanotoxicology Nanomed. 2017;2:11\u0026ndash;27. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.4018/JNN.2017010102\u003c/span\u003e\u003cspan address=\"10.4018/JNN.2017010102\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAl-Amin M, Dey S, Rashid T, Ashaduzzaman M, Shamsuddin S. Solar Assisted Photocatalytic Degradation of Reactive Azo Dyes in Presence of Anatase. Titanium Dioxide. 2016;2:14\u0026ndash;21. Accessed 25 Oct 2024.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eManzo M, Ahmed H, Nasrazadani S. Study on emission spectral lines of hematite and magnetite for purity\u0026rsquo;s differentiation. AIP Adv. 2020;10. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1063/1.5143413\u003c/span\u003e\u003cspan address=\"10.1063/1.5143413\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMysak E. PY083 diarylide AADMC yellow 275\u0026ndash;0570.TIF. CAMEO. 2020. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://cameo.mfa.org/wiki/File:PY083_diarylide_AADMC_yellow_275-0570.TIF\u003c/span\u003e\u003cspan address=\"https://cameo.mfa.org/wiki/File:PY083_diarylide_AADMC_yellow_275-0570.TIF\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Accessed 25 Oct 2024.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eQuinacridone. GuideChem. 2023. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.guidechem.com/encyclopedia/quinacridone-dic7717.html#:~:text=Quinacridone%2C\u003c/span\u003e\u003cspan address=\"https://www.guidechem.com/encyclopedia/quinacridone-dic7717.html#:~:text=Quinacridone%2C\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e with the chemical formula,paints%2C inks%2C and dyes. Accessed 25 Oct 2024.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePV 23 \u0026ndash; dioxazine purple. CHSOS 2024. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://chsopensource.org/pv-23-dioxazine-purple/\u003c/span\u003e\u003cspan address=\"https://chsopensource.org/pv-23-dioxazine-purple/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Accessed 25 Oct 2024.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDoke JGRABIT. MATLAB. 2024. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.mathworks.com/matlabcentral/fileexchange/7173-grabit\u003c/span\u003e\u003cspan address=\"https://www.mathworks.com/matlabcentral/fileexchange/7173-grabit\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Accessed 25 Oct 2024.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePY 139 \u0026ndash; isoindoline yellow. CHSOS 2024. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://chsopensource.org/py-139-isoindoline-yellow/\u003c/span\u003e\u003cspan address=\"https://chsopensource.org/py-139-isoindoline-yellow/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Accessed 25 Oct 2024.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZappielo CD, Nanicuacua DM, Santos WNL, dos, Silva DLF da, Dall\u0026rsquo;Ant\u0026ocirc;nia LH, de Oliveira FM, Clausen DN, Tarley CRT. Solid Phase Extraction to On-Line Preconcentrate Trace Cadmium Using Chemically Modified Nano-Carbon Black with 3-Mercaptopropyltrimethoxysilane. J Braz Chem Soc. 2016. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.5935/0103-5053.20160052\u003c/span\u003e\u003cspan address=\"10.5935/0103-5053.20160052\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNational Center for Biotechnology Information. PubChem Compound Summary for CID 24587, Barium permanganate 2024. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://pubchem.ncbi.nlm.nih.gov/compound/Barium-permanganate\u003c/span\u003e\u003cspan address=\"https://pubchem.ncbi.nlm.nih.gov/compound/Barium-permanganate\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Accessed 25 Oct 2024.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eToxicological Profile for Barium and Barium Compounds. Agency Toxic Subst Dis Regist 2007. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.ncbi.nlm.nih.gov/books/NBK598777/\u003c/span\u003e\u003cspan address=\"https://www.ncbi.nlm.nih.gov/books/NBK598777/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Accessed 25 Oct 2024.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSkocaj M, Filipic M, Petkovic J, Novak S. Titanium dioxide in our everyday life; is it safe? Radiol Oncol. 2011;45. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.2478/v10019-011-0037-0\u003c/span\u003e\u003cspan address=\"10.2478/v10019-011-0037-0\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRacovita AD. Titanium Dioxide: Structure, Impact, and Toxicity. Int J Environ Res Public Health. 2022;19:5681. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3390/ijerph19095681\u003c/span\u003e\u003cspan address=\"10.3390/ijerph19095681\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHolopainen M, Vallyathan V, Hedenborg M, Klockars M. Toxicity of Phlogopite and Muscovite. Vitro. Heal. Relat. Eff. Phyllosilicates. Berlin, Heidelberg: Springer Berlin Heidelberg; 1990. pp. 349\u0026ndash;60. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/978-3-642-75124-0_30\u003c/span\u003e\u003cspan address=\"10.1007/978-3-642-75124-0_30\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCai L, Wang J, Peng J, Tan Z, Zhan Z, Tan X, Chen Q. Characteristic of microplastics in the atmospheric fallout from Dongguan city, China: preliminary research and first evidence. Environ Sci Pollut Res. 2017;24:24928\u0026ndash;35. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s11356-017-0116-x\u003c/span\u003e\u003cspan address=\"10.1007/s11356-017-0116-x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDris R, Gasperi J, Saad M, Mirande C, Tassin B. Synthetic fibers in atmospheric fallout: A source of microplastics in the environment? Mar Pollut Bull. 2016;104:290\u0026ndash;3. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.marpolbul.2016.01.006\u003c/span\u003e\u003cspan address=\"10.1016/j.marpolbul.2016.01.006\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFeng J. Toxicological significance of azo dye metabolism by human intestinal microbiota. Front Biosci. 2012;E4:568. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.2741/400\u003c/span\u003e\u003cspan address=\"10.2741/400\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"},{"header":"Tables","content":"\u003cp\u003eTable 1 to 4 are available in the Supplementary Files section.\u003c/p\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"microplastics-and-nanoplastics","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"mina","sideBox":"Learn more about [Microplastics and Nanoplastics](http://microplastics.springeropen.com)","snPcode":"43591","submissionUrl":"https://submission.nature.com/new-submission/43591/3","title":"Microplastics and Nanoplastics","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-5334015/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-5334015/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eMicroplastic (MPL) abundance in the environment and the biosphere is a grave problem that is confounded by many aspects, one vital aspect being the characterization of their heterogeneous matrix. Currently, spectroscopy, chromatography, and soxhelation aid in this matter. However, many of these techniques are time consuming for MPL characterization, which can include a large number of particles. Therefore, we propose a facile \u0026ldquo;Additive Analysis\u0026rdquo; algorithm that can provide the top ten matches for additives for an MPL. For our first trial, we used 2 MPL entries, from FLOPP-E (C2. Blue Fiber) and SLOPP-E (Polyester 12. Red Fiber), as a continuation of our previous work. For our second trial, we extended the use of the algorithm to a semi-randomly selected subset of MPL samples from FLOPP-E and SLOPP-E based on choosing 1 sample of each color for each polymer. Both trials\u0026rsquo; reference used an in-lab digitization of the Hummel database for Fourier-transform Infrared (FTIR) spectroscopy and an open-source Raman spectroscopy database from Nava. We determined that the \u0026ldquo;C2. Blue Fiber\u0026rdquo; contains amounts of a metal-free phthalocyanine, potentially indicating the presence of degradation in context to the controls (t\u003csub\u003e10,.05\u003c/sub\u003e: .4879, p: .6387). For \u0026ldquo;Polyester 12. Blue Fiber,\u0026rdquo; we determined a high likelihood of significant amounts of quinone and azo-family colorants in the sample, negating a previous hypothesis of pyrrole presence (W: 0, p: .036364). For the second trial, 49/56 and 27/40 hits were generated out of the randomly selected samples, with a vast majority possessing hits (matching the color of the sample) within our most scrutinizing tolerance of 5 1/cm (77.6%/74.07%), respectively. For the FTIR portion, the top 3 IDs from tolerances of 5, 10, and 15 1/cm were benzenesulfonohydrazide (1st and 2nd Hit), titanium dioxide (4th Hit), and barium permanganate/barium sulfate (6th Hit). For the Raman portion, the top 3 IDs from tolerances of 5, 10, and 15 1/cm were PR210 (azo derivative \u0026ndash; 2nd Hit), PB25 (azo derivative \u0026ndash; 2nd Hit), and muscovite (mineral \u0026ndash; 1st Hit). Lastly, the distribution for these hits appears to identify organic colorants (FTIR) and azo-derivative colorants (Raman) most dominantly. Our discussion concludes with the potential toxicological impacts of these top 6 IDs.\u003c/p\u003e","manuscriptTitle":"Semi-quantitative Computational Analysis of Plastic Additives in a FLOPP-E and SLOPP-E Database Subset","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-11-11 14:10:55","doi":"10.21203/rs.3.rs-5334015/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-01-11T09:00:29+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-01-07T20:08:27+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"144016849799616952605847361473118697247","date":"2024-12-23T10:52:15+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2024-12-22T20:57:18+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"246735240317132283929576977862739654432","date":"2024-11-29T18:34:29+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2024-11-29T05:22:35+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2024-11-02T14:26:56+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2024-10-30T11:54:34+00:00","index":"","fulltext":""},{"type":"submitted","content":"Microplastics and Nanoplastics","date":"2024-10-25T17:25:55+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"microplastics-and-nanoplastics","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"mina","sideBox":"Learn more about [Microplastics and Nanoplastics](http://microplastics.springeropen.com)","snPcode":"43591","submissionUrl":"https://submission.nature.com/new-submission/43591/3","title":"Microplastics and Nanoplastics","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"66d55962-7881-48bc-9695-901fd0f4a132","owner":[],"postedDate":"November 11th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2025-03-10T20:15:22+00:00","versionOfRecord":{"articleIdentity":"rs-5334015","link":"https://doi.org/10.1186/s43591-025-00114-z","journal":{"identity":"microplastics-and-nanoplastics","isVorOnly":false,"title":"Microplastics and Nanoplastics"},"publishedOn":"2025-03-05 15:57:44","publishedOnDateReadable":"March 5th, 2025"},"versionCreatedAt":"2024-11-11 14:10:55","video":"","vorDoi":"10.1186/s43591-025-00114-z","vorDoiUrl":"https://doi.org/10.1186/s43591-025-00114-z","workflowStages":[]},"version":"v1","identity":"rs-5334015","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-5334015","identity":"rs-5334015","version":["v1"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}