# 2D Fourier Differentiation in Matlab

I am working in Matlab to compute partial derivatives and a Laplacian using Fourier Transforms. I prefer to take spectral derivatives, since it corresponds to multiplication by a wave number, however, I think my code is off since a large bulk of my data after computation gets sent to zero. Here is my code,

K = 16; KT = 2*K;
dx = 2*pi/KT;
x = 0:dx:2*pi-dx;

% build spectral derivatives and laplacian
Dxk = 1i*ones(KT,1)*[0:K -K+1:-1];
Dyk = 1i*[0:K -K+1:-1]'*ones(1,KT);

D2x = Dxk.^2; D2y = Dyk.^2;
Imat = eye(KT,KT);

% creates derivatives using a kronecker product
Dx = kron(Dxk,Imat); Dy = kron(Imat,Dyk);

% Lap = kron(D2x,Imat) + kron(Imat,D2y);

[xx,yy] = meshgrid(x,x);
f = fft2(sin(xx).*cos(yy));
f_x = Dx.*f(:);
f_x = real(ifft2(reshape(f_x,KT,KT)));

surf(f)


However, my derivatives end up looking like the following, Fourier partial wrt x. Thanks in advance for the help!