Dear Steven, Thanks for your suggestion before. As you mentioned, I really need to fix a frequency first, but the wavelength equals to 0.5um, and the lattice period is 11um( I am now repeating a result from PRL paper), which makes the frequency 22 in MPB, too big to find the root? On the other hand, I am not sure about, since I define kdir in 'find-k' as (kx,ky,0) to find k-magnitude(since it's 2D, kz should be 0 here, am I right?) does the k-magnitude equal to kz(q)? I have attached my code below, I have also attached the band structure picture they got in attachment. Their values of q completely have no comparison with the results got from MPB, which makes me more confused now. So could you kindly give me some further suggestion to solve this problem? I really appreciate your help.
Best wishes Yiling ======================================================================== ; simulate the band structure in paper PRL-91-213906-2003-Two-dimensional Optical lattice soliton ; Sinusoidal lattice (define pi (deg->rad 180.0)) ; Define a function of position p (in the lattice basis) which returns ; the material refractive index at that position. ; This is periodic, and also has inverse symmetry. ; RI modulation is directly related to the voltage applied to the photorefractive crytal, ; which is called as index potential in paper-----V(x,y) ; For sinusoidal, V(x,y)=-(V0/2)[sin(pi*x)^2+sin(pi*y)^2] ; and the RI modulation is related to the normalized potential depth via V0=(2*pi*n0*a/lamda0)^2*delta. ; Then RI modulation: Delta=-(lamda0/2/pi/n0/a)^2*V0/2*[sin(pi*x)^2+sin(pi*y)^2] (define n0 2.3) (define a 11) (define lamda0 0.5) (define-param V0 21.6) (define rad 3.5) (define r (/ rad a)) (define (sinusoidal-func p) (make dielectric (index (+ n0 (* -0.5 V0 (sqr (/ lamda0 2 pi n0 a)) (+ (sqr (sin (* pi (vector3-x p)))) (sqr (sin (* pi (vector3-y p)))))))))) ; The periodicity is 11um, and the radius is 3.5um for the dielectric-function. (set! geometry (list (make cylinder (center 0 0 0) (radius r) (height infinity) (material (make material-function (material-func sinusoidal-func)))))) (set! default-material (make dielectric (index n0))) (set! geometry-lattice (make lattice(size 1 1 no-size))) (set! resolution 64) ; the wavelength is 0.5um, according to the implicit periodicity, omega shoule be a/lamda=22(???? is it too big) (define-param omega 22) ;????it's too big to find the root (define-param band-min 1) (define-param band-max 5) (define-param kdir (vector3 0.5 0 0)) (define-param tol 1e-4) (define-param kmag-guess 0.5) (define-param kmag-min -1.0) (define-param kmag-max 1.0) ;(find-k p omega band-min band-max kdir tol kmag-guess kmag-min kmag-max) (find-k NO-PARITY omega band-min band-max kdir tol kmag-guess kmag-min kmag-max) ****************************************************************************8 MPB can certainly solve any problem of that form. However, you still need a frequency in order to define your mode; given K (the in-plane wavevector with components kx and ky) and q (the z wavevector, although have switched the sign convention so that q = -kz) there are infinitely many eigenmodes with different frequencies. Without a frequency your problem seems to be ill-posed. MPB solves the problem where, given (kx,ky,kz) you find the frequencies. Alternatively, with find-k you can give (kx,ky) and the frequency and then solve for kz (q). Steven ------------------------------ _______________________________________________ mpb-discuss mailing list mpb-discuss@ab-initio.mit.edu http://ab-initio.mit.edu/cgi-bin/mailman/listinfo/mpb-discuss End of mpb-discuss Digest, Vol 32, Issue 7 ******************************************
<<attachment: bandstructure.JPG>>
_______________________________________________ mpb-discuss mailing list mpb-discuss@ab-initio.mit.edu http://ab-initio.mit.edu/cgi-bin/mailman/listinfo/mpb-discuss