Well this is an interesting question that I will try to answer thoroughly. Fluorine, the diatomic molecule F2, is probably harmful and toxic for the body in sufficient quantities. However the context of this question is to do with fluorine attached to a molecule. All in all,
a fluorine in a molecule is unlikely to be toxic in general. I will discuss specific cases, namely where the fluorine alters the pharmacodynamics of the molecule which then renders it toxic in some way.
Point to consider:
C-F Bond Strength: The C-F bond is very strong, in fact the strongest by quite a bit of all the C-Halogen bonds. There are 3 reasons why this is so:
-C and F are in the same row of the periodic table, meaning their orbitals are of similar energy, thus overlap well to create a strong bond.
-F is the most electronegative element of the periodic table, and the difference in electronegativity between C and F is large, and the largest of all C-Halogen bonds, increasing the strength of the ionic component of the bond.
-F-, the fluoride ion, is a terrible leaving group; the small ionic radius of the fluoride ion means it is not able to delocalise the negative charge it gains very well.
As a result, F in an alkyl chain will rarely be displaced in a substitution reaction (with other halogens though, this is a possibility and needs to be discussed later), and F on an aromatic ring will rarely be displaced as well (aromatic rings do not in general undergo substitution reactions, but I will consider the exception below as well).
Thus F is sometimes thought of as similar to hydrogen, being small in size and relatively inert. The only difference is F can act as a hydrogen bond acceptor (HBA), allowing the molecule to form slightly different interactions with the protein it is interacting with. This sometimes is enough to completely change the molecule's binding profile, and we are going to look at amphetamine and 4-fluoroamphetamine as a case study.
What happens when you put a fluorine on amphetamine (in the para position on the ring)?
Amphetamine normally interacts with the dopamine transporter to release dopamine. When the molecule binds to DAT, its phenyl ring interacts with hydrophobic residues V120 annd F325 [1]. Once you put a fluorine in the para position of amphetamine, this not only reduces the binding interactions with DAT, but also greatly increases interaction with SERT (by accepting a hydrogen bond from Y176).
Thus 4-FA now releases serotonin, is neurotoxic and has MDMA like effects.
When should I not take a fluorine compound?
We now know that in general fluorine compounds, whether on aromatic rings or alkyl chains, are unlikely to be metabolically toxic, we have discovered that fluorine can often change the pharmacodynamics of the drug, which could in turn render it more or less toxic. I will now quickly look at halogen compounds that ARE toxic.
First we consider the mustard gasses (eg sulfur mustard:
https://en.wikipedia.org/wiki/Sulfur_mustard). Here we have 2 chlorines rather than fluorines, much better leaving groups, and also we have the critical nitrogen atom, which can displace the chlorines to form a reactive aziridinium ion intermediate, which can be opened up by nucleophiles. If the nucleophile is DNA, this can eventually lead to cancer. Note that the nitrogen makes this compound way more toxic than it would be without it, by the process I have just described:
anchimeric assistance.
Next, we consider aromatic rings where fluorine could be dangerous. Specifically, aromatic rings with electron withdrawing substituents ortho and para to the fluorine. This will promote a reaction called nucleophilic aromatic substitution, and if you are interested in this then you can read more about it in any organic chemistry text.
[1]
http://www.nature.com/nature/journal/v521/n7552/abs/nature14431.html
