The acetylene flow rate (45 L h−12−5) and the observation height (9 mm) were automatically optimized for maximum absorbance of NO at 215.360 nm, using the spectrometer software. The WSA was evaluated in the 1 (center pixel, CP) to 9 pixel (CP ± 4) interval, where CP
coincided with the line center. The main figures of merit (slope and linear correlation coefficient of the calibration curve, characteristic concentration,limit of detection, and relative standard deviation) are given in Table 1. The data revealed that when the integration area was altered from 1 to 9 pixels, the slope of the analytical curve increased approximately
6-fold and the characteristic concentration (C) decreased from 824.5 to 136.5 mg L−1. Analytical curves with satisfactory linearcorrelation coefficients were consistently obtained. The LOD decreasedfrom 170.3 to 75.4 mg L−1 o when the number of pixels was increased from 1 to 5. The spectrum for NO presents a broad peak at the 215.360 nm line (Fig. 2a), allowing the use of a greater integration area to improve the LOD. On the other hand, when the number of pixels
was increased from5 to 9, the LOD increased from75.4 to 176.5 mgL.The full width at half maximum(FWHM) provides a measure of the dispersion of the absorbance profile over pixels of an analytical line, as well as the magnitude of absorbance at adjacent pixels. The asymmetry of
the broad absorption line for NO at 215.360 nm suggests that the main contribution of WSA N5 pixels to the analytical signal was noise, which caused the LOD to deteriorate. An integration area equivalent to 5 pixels was selected for use in subsequent experiments because it provided the best LOD (75.4 mg L−1), RSD (2.6%), and linear regression correlation coefficient (r 0.9994). A typical calibration for N is shown in theSupplementary material (Fig. S1a).