Convert layers of 3D matrix to list of cells

I'd like to take a 3D matrix and convert it to a list (1D) of cells.

Example:

x(:, :, 1) = magic(3);
x(:, :, 2) = 2*magic(3);

I want c{1} to be x(:, :, 1)
I want c{2} to be x(:, :, 2)

My end goal is to create a block diagonal 2D matrix of the layers of x such that,

z = [x(:,:,1),        0
          0, x(:,:,2)]

or for complete...

z = [8, 1, 6,  0, 0,  0;
   3, 5, 7,  0, 0,  0;
   4, 9, 2,  0, 0,  0;
   0, 0, 0, 16, 2, 12;
   0, 0, 0, 6, 10, 14;
   0, 0, 0, 8, 18,  4]

I know I can do this very easily in a loop, but there has to be a better solution. I like the following solution but it assumes that every matrix along the diagonal is the same:

blk = magic(3);
Arguments = repmat( {blk}, 1, 100 );
A = blkdiag( Arguments{:} );

Thanks in advance

Answer 1

[Azzi Abdelmalek]

Edit

x(:, :, 1) = magic(3);
x(:, :, 2) = 2*magic(3);
x(:, :, 3) = 3*magic(3);
x=arrayfun(@(ii) x(:,:,ii),1:size(x,3),'un',0)
A = blkdiag( x{:} )

[Azzi Abdelmalek]

Stephen commented

Yes, that works for the simplified example I posted but what if x has 1000 layers? It gets pretty messy if size(x,3) is greater than just a few.

[Azzi Abdelmalek]

Look at edited answer

[Stephen]

Wow! That is quite nifty. I'm gonna have to read up on arrayfun. Thanks a lot for your help and if think of any other options, please add them to this page.

Also, is there an easy way to reverse this task? As in take A back into original x? (assign elements of block diagonal back to layers of a 3D matrix)