“…As shown in Figure F, the value of I 883 / I 998 for d -Pro is about 21-fold greater than that for l -Pro (after blank correction), while the value for d -Ala is about 20-fold greater than that for l -Ala, indicating an excellent enantiomeric selectivity and detection performance free from the interferences of l -Pro and l -Ala. What is more, other possible interferents in saliva were also checked such as KHCO 3 , CaCl 2 , NaCl, MgCl 2 , glutathione (GSH), glycine (Gly), cysteine (Cys), valine (Val), serine (Ser), threonine (Thr), peroxidase (Perid), amylase, vitamin B1 (VB1), lysozyme (LZ), phospholipase D (PLD), and α-glucosidase (α-glu). Again, in Figure F, the values of I 883 / I 998 for these species are much lower than those for d -Pro and d -Ala. As summarized in Table S1, the detection performance (linear range, LOD, and anti-interference) of the current SERS method is comparable with other fluorescence, electrochemical, colorimetric, and electrochemiluminescence methods for d -Ala and d -Pro. ,− Compared with these assays, our method avoids complex enzyme immobilization procedures and high-temperature (up to 900 °C) synthesis conditions in an inert atmosphere, and the saliva sample pretreatment is also simple.…”