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.

6C h a p t e r 2 Literature review To identify the range of spectroscopic methods applicable to the analysis of transportation infrastructure materials, the research team conducted a comprehensive literature search. The team explored bibliographic databases using the Trans- portation Research Information Service (TRIS) and then expanded the search to include various Internet bibliographic sources, such as the ASCE Civil Engineering Database, ScienceDirect, and others. This approach did not limit the sources to the fields of transportation and civil engineering but included various journals from the chemistry and petro- leum industries. All references were compiled in a computer- ized bibliographic database using Bibus software. The literature review was divided into the following topics: • Underlying principles of the most commonly used spec- troscopic methods for material analysis (i.e., infrared and Raman spectroscopy; gas, gel permeation, and liquid chro- matography; X-ray diffraction and fluorescence; and nuclear magnetic resonance). • Practical application of the above methods to identify sub- stances of interest (e.g., paints and coatings, antistripping agents, phosphoric acid additive, air oxidation, polymer additives, asphalt emulsions, water content, curing agents, and portland cement content). • Evaluation of portable spectroscopic devices and their applicability to on-site analysis of construction materials for availability, applicability, limitations, complexity of usage, sensitivity, and reliability. • Review of the existing ASTM/AASHTO standards relevant to spectroscopic testing of transportation construction materials. • On the basis of the findings, potentially promising spec- troscopic techniques were recommended for laboratory evaluation with emphasis on portable, yet cost-effective, instruments. The most important review findings are discussed in Chapter 3, while the complete literature review with references can be found in Appendix E. Surveys and Workshop To identify the needs for spectroscopic testing among the SHAs and to develop criteria for feasibility and imple- mentability of the field methods, the SHRP 2 coordinators from 50 states were surveyed. The survey questionnaire asked for the most challenging materials in terms of QC in the field, QA/QC procedures used by SHAs, and desirable features of those procedures and related instruments. In addition, a workshop with experts from both SHAs and the industry was held before finalizing the experimental program for Phase 2. The workshop concentrated on qualitative and quantitative requirements for field QA/QC procedures that would be implemented as a result of this project. The workshop addressed sample preparation, test duration, reliability, train- ing effort, and equipment price. On the basis of the workshop feedback, a list of construction materials and desired testing and equipment parameters was compiled. These parameters were used to rank the feasibility of the spectroscopic devices to be chosen for laboratory and field evaluation. An additional survey of the 50 SHAs was conducted before field implementation of the successful procedures (Phase 3). The survey confirmed the relevance of the compiled material list to the needs of material testing professionals and helped to finalize the list of materials for field testing. The main find- ings of the two surveys and the workshop are discussed in Chapter 3, and details are provided in Appendices F and G. Development of testing Matrix The testing matrix for “proof-of-concept” laboratory analyses was developed based on the literature review, SHA surveys, input from the workshop participants, and the experience of Methodology

7the research team. The matrix included portable FTIR, Raman, XRF, and X-ray diffraction (XRD) instruments, as well as sta- tionary (laboratory-based) infrared (IR), XRD, gel permeation chromatography (GPC), and nuclear magnetic resonance (NMR) equipment. This approach ensured that the capabili- ties of portable equipment would be compared with tradi- tional stationary equipment and methods that are currently immature for field use would be considered in the event of future technological advances. This proved especially true for methods with only few portable instruments available, such as XRD and NMR. To finalize the list of brands to be tested within each mate- rial category, the team conducted a survey of the approved and qualified product lists (APL/QPL) that were available on 34 out of 50 surveyed SHA websites. The top five materials (which appeared most often on APL/QPL) were identified for each group, and one or two were included in the test- ing matrix. The finalized test matrix and list of materials, along with the details of the material survey, are provided in Appendix A. Laboratory analyses Laboratory experiments were performed in Phase 2 to inves- tigate the applicability of each method to testing selected materials, as summarized in Appendix A, Table A.2. This table also summarizes the specific objectives for each material cat- egory and method combination. Two possible outcomes for spectroscopic testing were identified: (1) verification of the chemical composition (if provided by the manufacturer) or determination of the signature spectrum for pure materials and compounds and their components as supplied by manu- facturers; and (2) detection and, if possible, quantification of additives and contaminants in a material. In addition, Phase 2 had the following common objectives for all material categories: • Compare stationary and portable equipment in terms of detection limit, accuracy, and precision. • Develop libraries of standard spectra for each material in the testing matrix. • Identify the material–methods combinations that are most promising for field testing and provide justification for those that are not. • Develop generic laboratory procedures for the most prom- ising material–method combinations. On the basis of the results of laboratory experiments in Phase 2, the research team recommended the portable devices that were deemed most suitable for further evaluation in the field (Phase 3). The following criteria guided the selection process (more details on equipment and testing protocols can be found in Appendix A): • Successful fulfillment of a specific objective for a particu- lar combination of method and material, as stated in Table A.2; • Specific procedure parameters, such as minimum sample preparation, time, and labor effectiveness (details are pro- vided in Table A.4); • Specific equipment characteristics, such as accuracy, reli- ability, and duration of measurement (details are provided in Table A.4); and • Equipment portability (size and weight) and costs (details are provided in Table A.4). Field Verification The main goal of Phase 3 was to test portable equipment identified as successful in Phase 2 under field conditions and to determine whether its use can satisfy the desirable QA/QC criteria. A series of field trips to various construc- tion projects was performed to verify the applicability of laboratory-proven portable spectroscopic devices to the chosen material categories. On-site testing was performed by members of the research team as well as site personnel (field technicians) to evaluate the feasibility of selected devices and to receive feedback on the field procedures. The proposed generic field testing procedures (see Appendix B) were refined based on the experience at the construction sites and test results. Data analysis Methods Two types of data analysis were conducted depending on the material–method combination. They were qualitative analy- sis and quantitative analysis. For qualitative analysis, one or more characteristic fea- tures of a spectrum are identified (e.g., absorption peaks on an infrared or Raman spectrum) for each material. These characteristic features may then be used to fingerprint a material by matching the experimental spectrum to a typi- cal spectrum for that material that has been saved in a spec- tral library. The quality of the match can be evaluated visually or by using simple parameters such as a “hit quality index” (HQI) for infrared spectra (see Equation 2.1) (1). To compare HQI values of a sample spectrum with a reference spectrum, their spectra were normalized as shown in Equa- tion 2.2 to produce spectral response values (absorbance) between 0 and 1 (1).

8HQI A L A L i ii n ii n ii n = −       = = = ∑ ∑ ∑ 1 2 11 2 1 2 1 ( . ) where Ai = normalized absorbance of the unknown spectrum at ith wavelength (or wave number) and Li = absorbance of the unknown spectrum at ith wavelength (or wave number). A A A A A i i = ′− − min max min ( . )2 2 where A′i is the measured absorbance of the infrared incident beam. Note that the vast majority of infrared spectrometers are equipped with customizable software that can automate the HQI calculation and qualitative match to the library spectra. The library can be provided by a vendor or created experimentally by the user. Similar approaches are adopted for qualitative analysis of XRD and Raman spectra. Fingerprinting of pure materials using XRF was accom- plished using quantitative criteria, given that XRF produces the quantitative elemental composition of the analyzed material. In this case, the concentration of critical elements in the material tested in the field (e.g., Ti in paints) can be used as a criterion to match the expected concentration in the pure material that has been established in the lab beforehand. Special chemometric analysis of the IR spectra was con- ducted to quantify the composition of complex samples, such as polymer-modified asphalt or a diluted epoxy coating system. Calibration curves were developed by measuring pure samples spiked with incremental and known concen- trations of an additive or contaminant (e.g., polymer in asphalt or water in diesel fuel). This approach is based on the Beer-Lambert law, according to which the concentration C of a component is directly proportional to its IR absorbance A at constant path length l, as in Equation 2.3 (1): C A el = ( . )2 3 where e is the molar extinction coefficient, which is constant for a particular component but usually unknown for com- plex materials. Finally, certain analyses involved the observation of a rela- tive change in the IR spectra attributable to a physical process, rather than change in composition. Such cases are the deter- mination of oxidation in recycled asphalt pavements (RAP) and the identification of chemical admixtures in PCC. To track spectral changes, the research team employed semi- quantitative analysis based on peak-to-peak ratio or valley- to-valley band integration approaches. More details on the semiquantitative approach to spectral analysis are provided in Chapter 3, Appendix N, Appendix O, and elsewhere (2). repeatability and reproducibility of test results The repeatability and reproducibility of test results on the most successful material–method combinations were evalu- ated in both laboratory and field phases of the project. The evaluation of repeatability concerned the variation in mea- surements taken by a single operator or instrument on the same item and under the same conditions. The reproduc- ibility of test results was judged by the level of variability in the results measured by independent operators on the same equipment. It should be noted that because of schedule con- straints, only a limited number of materials were included in the repeatability and reproducibility study, as summarized in Chapter 3 and detailed in Appendix P. Finally, the spectra or quantities obtained in the field for a specific material were compared with the spectra or quantities measured in the labo- ratory. This comparison facilitated conclusions on precision and accuracy of a particular material–method combination. project Deliverables Spectral Library A library of reference spectra for pure materials analyzed in Phase 2 was created that can be used for fingerprinting of these materials in the field. The database consists of a series of files with fingerprinting number values and key to the material labels. It will supplement the standards developed under this project and can be used by QA/QC specialists. The electronic copy of the spectral library is located at www.trb .org/Main/Blurbs/167279.aspx. Provisional AASHTO Standards Generic testing procedures with sampling and data process- ing guidelines were developed for the attenuated total reflec- tance (ATR) FTIR, Raman, and XRF analysis of liquid and solid samples (see Appendix B). On the basis of Phase 3, field verification, the most successful spectroscopic testing proce- dures were expanded to draft AASHTO standard specifica- tions. These specifications cover the analysis of titanium content in traffic paints by XRF and identification of chemical admixtures by ATR (see Appendix C). The target audiences

9for these AASHTO standards are QA/QC personnel at the research and material divisions of SHAs. Field Operation Manuals Field operation manuals were developed for the most success- ful spectroscopic devices (i.e., ATR and XRF instruments). These manuals target technical personnel designated by a transportation agency for the spectroscopic testing. They are expected to facilitate understanding of the somewhat sophisti- cated documents provided with the spectroscopic devices that are usually designated for a researcher. The field manuals are included in Appendix D. references 1. Duckworth, J. H. Spectroscopic Quantitative Analysis. In Applied Spectroscopy: A Compact Reference for Practitioners, Academic Press, Chestnut Hill, Mass., 1998, pp. 93–107. 2. Yut, I., and A. Zofka. Attenuated Total Reflection Fourier Transform Infrared Spectroscopy of Oxidized Polymer-Modified Bitumens. Applied Spectroscopy, Vol. 65, No. 7, 2011, pp. 765–770.