Data for: "Detection of multiple nitroaromatic explosives via formation of a Janowsky complex and SERS"

The purpose of this dataset is to show that SERS detection of multiple nitroaromatic explosives (TNT, tetryl and HNS) can be achieved through formation of a Janowsky complex using 3-mercapto-2-butanone as the starting material.
Abbreviations
3M2B – 3-mercapto-2-butanone
DBU - 1,8-diazobicyclo (5.4.0) undec-7-ene
TNT – 2,4,6-trinitrotoluene
Tetryl – 2,4,6-trinitrophenylmethylnitramine
HNS – hexanitrostilbene
SERS – surface enhanced Raman scattering
PCA – principal component analysis

Figure 2 – Absorbance spectra were collected using an Agilent Cary60 UV-Vis Spectrophotometer. Files were saved in .csv format and converted to an Excel spreadsheet (.xlsm) for processing. Each spectrum was plotted into a graph using Excel which produced the spectra shown in figure 2.

Figure 3 – (a) Absorbance spectra were collected using an Agilent Cary60 UV-Vis Spectrophotometer. Files were saved in .csv format and converted to an Excel spreadsheet (.xlsm) for processing. One spectrum was recorded every minute for 30 minutes. Each spectrum was plotted (overlayed) into a graph using Excel which produced the spectra shown in figure 3. (b) The absorbance at 475 nm was plotted as a function of time for the first 5 minutes. The resulting graph produced figure 3 (b).

Figure 4 – SERS spectra were collected using a Renishaw Plate Reader with 532 nm excitation. 3 replicate samples of both the control and sample which contained TNT were analysed and averaged. The data was collected as WiRE 2.2 (.wxd) files and converted to .spc files for processing in MatLab 2016. The spectra were truncated between 600 cm-1 and 1600 cm-1 and plotted overlayed. This resulted in the spectra shown in figure 4.

Figure 5 – (a) SERS spectra were collected using a Renishaw Plate Reader with 532 nm excitation. Samples containing a final TNT concentration of 0.11 µM, 0.55 µM, 1.11 µM, 1.66 µM and 2.22 µM were analysed (3 replicate samples, 5 measurements) as well as a control sample containing no TNT (3 replicate samples, 5 measurements). The data was collected as WiRE 2.2 (.wxd) files and converted to .spc files for processing in MatLab 2016. The spectra were truncated between 600 cm-1 and 1600 cm-1, baseline corrected and plotted overlayed. This resulted in the spectra shown in figure 5 (a). Peak intensity at 1300 cm-1 was calculated by subtracting the intensity at 1257 cm-1 from the intensity at 1300 cm-1. This value was then plotted as a function of TNT concentration using Excel, which resulted in the graph shown in figure 5 (b). Error bars represent ± 1 standard deviation. The limit of detection was calculated using 3 x standard deviation of the control sample / the gradient of the line.


Figure 6 - SERS spectra were collected using a Renishaw Plate Reader with 532 nm excitation. 3 replicate samples of both the control and sample which contained TNT, tetryl and HNS were analysed and averaged. The data was collected as WiRE 2.2 (.wxd) files and converted to .spc files for processing in MatLab 2016. The spectra were truncated between 600 cm-1 and 1600 cm-1 and plotted offset for clarity. This resulted in the spectra shown in figure 6 (b).

Figure 7 - (a) SERS spectra were collected using a Renishaw Plate Reader with 532 nm excitation. Samples containing a final Tetryl concentration of 0.11 µM, 0.55 µM, 1.11 µM, 1.66 µM and 2.22 µM were analysed (3 replicate samples, 5 measurements) as well as a control sample containing no Tetryl (3 replicate samples, 5 measurements). The data was collected as WiRE 2.2 (.wxd) files and converted to .spc files for processing in MatLab 2016. The spectra were truncated between 600 cm-1 and 1600 cm-1, baseline corrected and plotted overlayed. This resulted in the spectra shown in figure 5 (a). Peak intensity at 1329 cm-1 was calculated by subtracting the intensity at 1300 cm-1 from the intensity at 1329 cm-1. This value was then plotted as a function of Tetryl concentration using Excel, which resulted in the graph shown in figure 7 (a) right. Error bars represent ± 1 standard deviation. The limit of detection was calculated using 3 x standard deviation of the control sample / the gradient of the line.
(b) SERS spectra were collected using a Renishaw Plate Reader with 532 nm excitation. Samples containing a final HNS concentration of 1.1 µM, 2.3 µM, 4.5 µM, 6.7 µM and 8.9 µM were analysed (3 replicate samples, 5 measurements) as well as a control sample containing no HNS (3 replicate samples, 5 measurements). The data was collected as WiRE 2.2 (.wxd) files and converted to .spc files for processing in MatLab 2016. The spectra were truncated between 600 cm-1 and 1600 cm-1, baseline corrected and plotted overlayed. This resulted in the spectra shown in figure 5 (a). Peak intensity at 1336 cm-1 was calculated by subtracting the intensity at 1305 cm-1 from the intensity at 1336 cm-1. This value was then plotted as a function of HNS concentration using Excel, which resulted in the graph shown in figure 7 (a) right. Error bars represent ± 1 standard deviation. The limit of detection was calculated using 3 x standard deviation of the control sample / the gradient of the line.

Figure 8 - SERS spectra were collected using a Renishaw Plate Reader with 532 nm excitation. Spectra were recorded of samples containing all three explosives, TNT & Tetryl, TNT & HNS, Tetryl and HNS and each explosive complex individually (3 replicate samples, 5 measurements). The data was collected as WiRE 2.2 (.wxd) files and converted to .spc files for processing in MatLab 2016. The spectra were truncated between 600 cm-1 and 1600 cm-1, baseline corrected and plotted offset. PCA was performed on these datasets using MatLab 2016, producing the principal loadings and corresponding scores plot shown in figure 8.

Supplementary Information

Figure S1 - SERS spectra were collected using a Renishaw Plate Reader with 532 nm excitation. Samples were prepared which contained no TNT to act as control samples. The data was collected as WiRE 2.2 (.wxd) files and converted to .spc files for processing in MatLab 2016. The spectra were truncated between 600 cm-1 and 1600 cm-1, baseline corrected and plotted overlayed, resulting in the spectra shown in figure S1.

Figure S2 - Absorbance spectra were collected using an Agilent Cary60 UV-Vis Spectrophotometer. Files were saved in .csv format and converted to an Excel spreadsheet (.xlsm) for processing. Each spectrum was plotted into a graph using Excel which produced the spectra shown in figure 2.

Figure S3 - SERS spectra were collected using a Renishaw Plate Reader with 532 nm excitation. Spectra were recorded of samples containing all three explosives, TNT & Tetryl, TNT & HNS, Tetryl and HNS and each explosive complex individually (3 replicate samples, 5 measurements). The data was collected as WiRE 2.2 (.wxd) files and converted to .spc files for processing in MatLab 2016. The spectra were truncated between 600 cm-1 and 1600 cm-1, baseline corrected and plotted offset, resulting in the spectra shown in figure S3.