raja,

You will want to separate the resonance of the capacitor by increasing the resonance frequency.  To increase the resonance frequency of the cap you could make the interdigitated cap fingers shorter and add more in parallel to return to the same cap value.  

In general the oscillator will resonate at a frequency where the gain is the highest.  Gain will be construed as the gm*Re(Z(tank)).  I am not sure which frequency will be the oscillation frequency if they two frequencies have roughly equal gain, possibly some intermediate one.

http://rfcooltools.com

Raja,

Consider two WEAKLY coupled tanks one resonant tank1 at freq1 and one tank2 resonant at freq2.  Now consider the Q of the tank1 to be greater than tank2.  Apply a step response  and observe the voltage waveforms after a length of time. the tank with the higher Q will die down slower than the tank with the lower Q also note that the resonant frequency might be a combination at first, but later each tank will settle to a their own frequency separately.   (this is not to say that the final frequencies of both aren't altered by the weak coupling element).  

If you are with me so far...

Now consider this step response in reverse as energy is added to the tanks.  The tank with the higher Q will increase amplitude at a faster rate.  


Finally, the active part of the VCO will eventually become a limiter as the amplitude further increases.  Now that the VCO is a full limiter the selectivity of that tanks resonant frequency is reinforced since the additive  frequency of the weaker tank is only present on the through zero transitions (or when the limiter changes polarity).  The stronger resonance is further reinforced by the periodic energy additions from the limiting action and eventually very little energy is at the secondary resonance and thus a single resonance oscillation occurs for the most part.  

http://rfcooltools.com