So which one are you trying to do: tune a PID controller or design a controller using pole placement. Pole placement is done on a full state feedback, PID would act on just the system output.
To design a PID controller:
Tf = (-1.41)/(s^2-1.5*s-2.25); %define plant transfer function
C=pidtune(Tf,'pidf') % design a PID controller with filter derivative
step(feedback(C*Tf,1)); % compute and plot the step response
To design pole placement controller:
sys=ss(Tf); % convert to state-space form
K=place(sys.A,sys.B,[-1 -2]); % design pole placement full state feedback gain to place poles at -1 and -2
