Several things. First:
for i = length(T_K)
should be
for i = 1:length(T_K)
The first way just iterates once, using the last element of T_K, which is why the first four elements of p1_pure are zero.
Second: T_K in the code is in Kelvin, but T in the formula is in Celsius, so you would have to use T_K-273.15 in your MATLAB calculation to convert it.
Third: exp() is natural (base e) exponentiation, but the formula says log10 (not natural log), so your MATLAB code would have 10^() rather than exp().
Making those changes still doesn't give a result that matches what Excel gives you, so I would check the units of a, b and c, and check your Excel formula.
a = 4.00266;
b = 1171.53;
c = -48.784;
T_K = 300:10:340;
for i = 1:length(T_K)
p1_pure(i) = 10^(a-b/(T_K(i)-273.15+c)); % make this a function if you want
end
p1_pure
