An artist named Rose-Lynn Fisher has made photos of human tear drops under a microscope, and shows them to be different in their structure depending on how those tears were generated. Sad tears have a different appearance on a molecular level than happy tears, and so on.
Unlike the notorious work of Masaru Emoto, who claimed that people's attitude towards water changed that water's molecular structure, the ideas of tears being different seems at least partially plausible. A person who was sad might have different hormones or levels of adrenalin, for example, than a person who was happy. Could some internal difference in the person crying find its way into the tears in any way that was observable?
However, even though the artist claims to have used "tools of science", it also seems equally likely that the artist has simply chosen from a wide sample of random images available from each set of tears and selected the ones that she felt matched her narrative.
As art, it's nice enough. As science... is there anything to it? Could tears be physically different depending on the reason the human produced them? Is there any research the demonstrates what the differences are and why?