3. Results
3.4 SERS analysis of APIs on cysteamine-based chemical sensors
3.4.2 Statistical evaluation of Tenofovir on glass and cysteamine-based chemical sensors
Linear curves of TDF on glass, Au@Cys and Ag@Cys substrates were constructed and compared using peak area versus concentration. The graphs were used to determine the R2 values and the calibration sensitivity as the slope of the line. Figure 31 below is three graphs of the 727 cm-1 adenine ring breathing mode.
97 Figure 3.31:Linear regression analysis of Tenofovir using the adenine ring functionality (723 cm-1). Peak area vs concentration on Glass (Top), Au@Cys/Au (left) and Ag@Cys/Ag (right).
Figure 3.31 above gives the statistical outlook of TDF Raman response on the different platforms, using the adenine peak area as a signal. The glass produced a R2 value of 0.86 (p<0.10) compared to the 0.96 (p<0.01) and 0.97 (p<0.01) values of the gold and silver samples respectively. In addition, the silver sensor showed the highest calibration sensitivity value of 13803, followed by the gold sensor at 5152 and the plain glass gave the lowest value of 670. The explanation for the observed statistical data can be addressed from the Raman shifts which show the most blue shift on the glass as the ring is more compressed, while the metal nanomaterial sensors show ring strain caused by molecular interactions that led to a relative red shift. This progression in values also shows that the metals played role in improving the linearity of the TDF signals by increasing the signal-to-noise response where the glass interference is reduced gradually as the thickness increases [53-54]. Furthermore, the improved calibration sensitivity speaks to the ability of the metals to improve the signal when the molecules are in the hotspot, indicated by Raman shifts mediated by intermolecular forces. The R2 and calibration values obtained in these experiments were used to calculate the other parameters to assess the efficiency of the sensors. Figure 32 below is an evaluation of the 1310 cm-1 peak area which is attributed to the combined contribution of the imine and the primary amine stretching modes.
98 Figure 3.32:Linear regression analysis of Tenofovir using the adenine ring functionality C=N, C-N, C-H functionality (1326 cm-1). Peak area vs concentration on Glass (Top), Au@Cys/Au (left) and Ag@Cys/Ag (right).
The data in the figure above shows the statistical response from the 1330 cm-1 vibrational band of the primary amide of TDF. The R2 of the glass, gave the lowest value of 0.90 (p< 0.05) while the gold value was 0.95 (p< 0.05) and the highest value was from the silver substrate at 0.97 (p<0.01).
The calibration sensitivity from this set of experiments resulted in the glass substrate producing the lowest output of 736.2 followed by the gold sensor at 6635 and silver again with the highest value of 27348. The trend in this experimental group correlates with the peak shifting observed where the red-shifted glass value implies that the TDF ring functional groups are not compressed individually compared to the ring itself, which is blue shifting due to bond length shortening [55- 57]. In addition, the metallic sensors showed consistency in blue shift because the functional group is compressed in stretching mode possibly by intermolecular forces. The results obtained from this statistical analysis were used to determine values for additional parameters that characterize the performance of chemical sensors. Table 12 below is a summary of statistical values obtained for TDF on the three substrates.
99 Table 3.12:Statistical values of TDF on glass, Au@Cys/Au and Ag@Cys/Ag.
According to the table above, the chemical sensors showed a trend in analytical sensitivity towards the adenine ring as follows: the glass gave the highest value of 39.26 followed by silver with 26.36 and lastly gold produced a low value of 16. This trend can arise by considering the standard deviation of the data from glass decreased relative to the slope or calibration sensitivity whose linearity is confirmed by the R2 value. Since the latter was low in magnitude, implying a poor linear fitting, the analytical sensitivity value received is not considered reliable. When comparing the nanomaterial sensors, silver surpassed the gold value due to the high calibration sensitivity and lower standard deviation. The limit of detection from glass was found to be 0.05 mg/ml, while lower values were obtained from gold at 0.01 mg/ml and silver and 0.003 mg/ml respectively. In the case of the amine bonds of the ring at 1310 cm-1, the analytical sensitivity of glass was higher than gold at 43.13 and 20.6 respectively, while the silver sensor held the highest value at 52.24. For this functional group, the R2 value on glass improved from the adenine ring.
However, it remained lower than the chemical sensors with an additional lower significance value and thus it was regarded unreliable. The limit of detection significantly dropped on the chemical sensors compared to the glass substrate, with silver producing the lowest value of 0.001 mg/ml and gold at 0.01 mg/ml. The overall observation from the table above is that silver produced the most optimum combination of parameter values against gold and glass. In the next section, experiments with Lamivudine were conducted to compare the performance of the chemical sensors against the glass.