The DFT and inverse DFT are mappings from a N-dimensional complex vector space to an N-dimensional complex vector space.
When you change N as you do by padding, then they are different, so you cannot expect the vectors to be the same.
What you can do is this:
spectreFFT2 = fft(y,pts);
y2 = ifft(spectreFFT2);
y2 = y2(1:length(y));
figure (2);plot(t,y,'b',t,y2,'r');
Since padding is just appending zeros, you simply remove the zeros at the end.
