Evaluating Applications of Field Spectroscopy Devices to Fingerprint Commonly Used Construction Materials (2013)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

206 A p p e n d i x n peak-to-peak Ratio Quantification of SBS in Polymer-Modified Asphalt Binders Three replicate binder–styrene–butadiene–styrene (SBS) mixtures for each of the SBS concentrations (1%, 3%, and 6%) were prepared over 2 months and scanned by the Bruker attenuated total reflectance (ATR) Fourier transform infrared (FTIR) spectrometer. The calibration equation for the quan- tification of SBS content in asphalt binder was developed based on the peak-to-peak ratio approach. Such an approach considered individual ratios of an absorption peak at either 966 cm-1 (A966) or at 700 cm-1 (A700) to the major peaks for which absorbance value did not change (e.g., A2920, A2850, A1460, and A1380 cm-1). The A966 and A700 absorption val- ues are respectively associated with polybutadiene and poly- styrene. Because the ratio apparently changed with increasing SBS content, it was possible to choose the best suited ratio value based either on 966 cm-1 or 700 cm-1. To achieve the best fit, the individual ratios of the two SBS-associated peaks were regressed on SBS content as in the example below: A A N966 2920 10 1= + +b b C e ( . ) where b0, b1 = regression coefficients, C = polymer content, and e = error. In the first stage of the regression analysis, it was found that higher than 3 wt% polymer modification of binder resulted in significant deviation from linear trend expected by the Beer-Lambert law, as shown in Figure N.1. In addition, the variability in ATR measurements increased dramatically with change in SBS concentration from 3 wt% to 6 wt%. Therefore, 6 wt% SBS-modified samples were excluded from the analy- sis. When zero to 3 wt% modification was considered and outliers were removed from the data set, a similarly high goodness of fit was respectively achieved for both polystyrene and polybutadiene by using peak ratios A700/A2920 and A966/A2920 as predictors for polymer content C. Figure N.2 shows calibration trends along with trendline equations, goodness of fit (R2), and standard error (SE). Based on the R2 values, it is suggested that the peak-to-peak ratio is a viable approach to quantification of styrene– butadiene polymers in asphalt binders. The main reason for variability in the ATR measurements is assumed to be non- uniformity of polymer-modified binder samples because of the differences in mixing procedures between replicate batches (e.g., time and temperature). Valley-to-Valley Band integration and normalization Oxidation of Polymer-Modified Binders and RAP To quantify oxidation-related changes in the infrared (IR) absorption, band areas were computed using the following systematic procedure: • Step 1. Collect the spectra as absorbance values (A) versus wave number (v˜). An absorbance value at each wavelength is calculated by dividing the measured absorbance value by the wavelength. • Step 2. Apply atmospheric and water compensation and rubber-band baseline correction using default Bruker OPUS 6.5 software settings. The software calculates base- line as a frequency polygon, consisting of 64 baseline points. The “rubber band” is stretched between the spectrum end points, as it follows the spectrum minima (Figure N.3). • Step 3. Smooth the spectrum line using the Savitzky-Golay algorithm (using a 10-point averaging interval and a third- order polynomial). Quantitative Analysis of ATR FTIR Spectra

207 Figure N.1. Regression plots for peak ratios with all samples included. Figure N.2. Calibration equations for SBS concentration based on peak-to-peak ratios. Figure N.3. Illustration of the rubber-band baseline correction.

208 • Step 4. Extract absorption peaks using a second-derivative approach. • Step 5. Assign the extracted peaks to chemical functional- ities in the bitumen sample (Figure N.4). • Step 6. Compute band areas (ARv˜) using valley-to-valley integration (Figure N.4). The above procedure was automated using computational code developed in MATLAB 7.6.0 environment. The ARv˜ values were normalized to the total sum of all band areas (SARv˜), and the indices24 for the aromatic, oxygen- containing, and polymer-related absorption bands were com- puted as follows: • Aromaticity Index IAR = AR1600/SARv˜ • Carbonyl Index ICO = AR1700/SARv˜ • Sulfoxide Index ISO = AR1030/SARv˜ • Hydroxyl Index IOH = AR3400–3100/SARv˜ Figure N.4. Identification of bitumen components on FTIR spectrum in (a) region between 3,800 and 2,600 cm1 and (b) fingerprinting region between 1,800 and 600 cm1. (a) (b) • Polybutadiene Index IPB = AR968/SARv˜ • Polystyrene Index IPS = AR700/SARv˜ The changes in the above indices over the aging cycle served as indicators of the oxidation rate and polymer degra- dation, as discussed in the following text. The change in polybutadiene and polystyrene index values served as an indicator of the change in polymer content dur- ing accelerated aging by rolling thin-film oven and pressure aging vessel procedures. No statistically significant changes in polymer-related indices (IPS and IPB for polystyrene and poly- butadiene content, respectively) throughout the aging cycle were found in this study (see Figure N.5). Note that the poly- mer content ranged between zero and 3.3 wt%. Such an out- come appears to be in line with the previous studies that reported swelling or degradation of SB-based polymers at concentrations higher than 4 wt%. Figure N.5. Comparison of IPB and IPS between aged binders modified with 3 wt% of SB polymers.