as I have the value "0" in some the filed these formulas didnt work. So I thought adding the value of "1" with the publication in each year (of each country) would be harmless!
I don't know that I would agree with this. It seems a gross misrepresentation of the data to substitute 1 where there are 0's. It seems to me that those 0's mean "no publications", and substituting "one publication" for those years is not correct. It also seems to me that we don't really have a good idea how to regress or represent the data, which makes it difficult, IMO, to even talk about how to put an unkown computation into a spreadsheet.

One question -- you use the term "growth rate" which tends to imply an exponential (y=A*exp(Bx)) type function. Are you certain that an exponential function is the most appropriate function to use here? Would other functions be more appropriate? Excel's regression functions (LINEST(), LOGEST(), GROWTH(), TREND(), etc.) can handle a wide variety of function types. The hardest part, sometimes, is deciding what function seems most correct for a given problem.