Have you looked at the total energy flow of your problem to find out just how many calories (or BTUs, or whatever your favorite unit is) you are trying to generate and what time scale you need to generate it on?
For instance, you want your reaction to generate a temperature of 150C. I can do that by spraying a solution of t-butyllithium through the air. Instant ignition, much higher than 150C. How much of what material are you actually planning on heating? What is the heat capacity of that substance?
Then you want it to maintain it's temperature for 2 minutes - once again, I can do that by pouring it into a vacuum flask. Apparently, however, you want this heat to be transferred into another object. In that case, it isn't the temperature of the object you should be concerned about, but the energy flux from the object. How much energy are you trying to transfer across what surface area, over what amount of time?
I could see the potential for a two-part reaction system in your process - one that generates a lot of heat very rapidly to bring your temperature up quickly, and another that generates heat more slowly, to maintain the temperature by adding energy at the same rate that energy is being radiated from your object. But the actual energy flows that you desire would be necessary before making any predictions about what reactions might be useful.