Cu and Zn readily binds to carboxylate groups, as Cu2 paddlewheel and Zn4O units. You will need a tripeptide with (at least) two free carboxylate groups for coordination. However, I think unless your peptide is very rigid, the structure will be neither crystalline nor stable.

Currently I'm working with a tripeptide with two carboxylates Is that rigid enough or should I change to another peptide?

And my question was rather about the conditions of temperature, pH and the rest

Mr. Ruiz,

There is still uncleasr the effect of MLCT and LMCT on coordination ability, optical and NLO properties. There is possible to have both the effects even in metal-organics with d10 configuration of the metal ion 

The synthesis of extended 3D structures in this fashion still relies on a lot of trial and error I'm afraid. What kind of solvents are you using? If you are aiming for a specifific SBU like Cu-paddlewheel/ Zn4O like mentioned above you might consider the literature for conditions in which those have been known to form as the coordination of single metal ions is not going to do it in this case, I think. Furthermore, if you are working with a tripeptide with two terminal carboxylate groups for instance, you essentially have a linear 2-fold node which would really only give you 2D sheets in case of paddlewheels, 3D would only be possible with Zn4O. Also, the remaining N-Donor sites in the peptide might interact with Zn/Cu as well as these are rather soft metals HSAB wise.

You might want to have a look at this article and its references:

Article Synthesis, Structure, Topology and Catalytic Application of ...

Stella Helten: Right now the solvents I've been using are DMF, DEF and distilled water with TRIS buffer adjusted to pH: 8.2. The tripeptide I'm working with has got 2 carboxylate groups as well as the amino groups and a thiol group. On the other hand, How could the source of the ion afect to the coordination? Right now I've only been capable of syntethising the Cu paddlwheel using Copper Acetate as the source. but no other paddlewheel is formed using other metal sources. Also Iom interested on how did you predict if the structure was more likely to be 2D with Cu and 3D only with Zn4O.

DMF and DEF are reasonable solvents for MOF synthesis obviously. What is the reason for the TRIS buffer? If it is due to solubility, maybe the application of the sodium salt of the peptide might be something to consider as aqueous basic additives might be detrimental to MOF formation, especially if you should aim for the Zn4O SBU. Influence of the anion originating from the metal precursor is often observed in MOF chemistry. Roughly speaking: The anion, due to its polarity and form, can influence the interaction between the molecules one aims to assemble. For a more detailed discussion I can recommend the article by Kaptejn et al. and references therein: CrystEngComm, 2013, 15, 9249-9257. As for the dimensionality of the networks: One considers the points of extension of the linker/the SBU, most commonly the carboxylate C-atoms. In the case of paddlewheels, that gives you a square-planar node, which is 2D. Now if your linker is bound to that kind of SBU at 2 points (linear or bent), you will remain with a 2D structure. There is a number of articles by Michael O'Keeffe explaining the characterisation of MOFs by nets which can be helpful for MOF design. However, if you have been successful with paddlewheels, maybe axial pillaring of 2D sheets via, for instance, 4,4'-bipyridine might be a good option.