imageColin’s latest idea:

Actually, what this experiment shows is that a more intense image is obtained on the impregnated linen [(with lemon juice)] than the control, untreated linen, for heated metal templates at the same temperature. You will have to take my word for now that images can be obtained at lower temperatures on the impregnated linen with scarcely if any image on untreated linen.

It may well be that none of the imaging on treated linen is due to modification of linen carbohydrates, whether the cellulose or even the more heat-sensitive hemicelluloses etc of the primary cell wall. It may be that ALL the image represents pyrolysed products in the lemon juice. In other words, one can have thermal imprinting using a heat-sensitive  invisible “ink” that need not damage or affect the linen itself. This makes it more probable that the image thickness meets that 2o0nm criterion!

Is Colin now agreeing to some of the criteria? I like the fact that he is still trying to find a solution that fits his worldview.