## Results and discussion Fig xA Typical results

The model predicts the order of magnitude of fv along the flame centerline with an acceptable accuracy as shown in Fig. 3. However, predicted centerline fv profile for B100 surrogate starts from z = 40 mm while the experiment does not detect any soot for z ? 50 mm. On the centerline, the model predicts WP1066 gradual increase for fv starting from z ≈ 40 mm to fv = 0.8 ppm at z = 60 mm, followed by a gradual decrease to zero in higher flame heights. However, the experiment on the centerline does not detect any soot for z ? 50 mm then fv jumps to fv = 0.7 ppm at z = 60 mm and gradually increases to fv = 1.8 ppm at z = 70 mm and then decreases to zero in higher flame heights. Thus, on the centerline the model seems to over-predict fv below z = 60 mm and under-predict fv above z = 60 mm. Both the model and the experiment show Archaea fv levels start to decrease from z ≈ 70 mm.