Finding the finite zero point of transfer function under different sampling periods

Given a transfer function
stay T=[0.0001,0.1] The value in the range is , Sampling uses a zero order holder , Find finite zero .

num=[1];
den=conv([1,1],conv([1,10],[1 100]));
T=logspace(-6,-1,500);
numd=rand(500,4);
dend=rand(500,4);
z=rand(2,500);
for i=1:1:500
    [numd(i,:) dend(i,:)]=c2dm(num,den,T(i),'zoh');
    z(:,i)=roots(numd(i,:));
end
semilogx(T,z(1,:),'k-');
hold on
semilogx(T,z(2,:),'r-');
text(10^(-6),z(1,1),num2str(z(1,1)))
text(10^(-6),z(2,1),num2str(z(2,1)))

result ：