Can anyone suggest me a better method for digitonin based thylakoid membrane solubilization?

I tried couple of time but the Chl a/b ratio was not as expected according to the previous reports. When I compared the Chl a/b ratio of all three fractions, I got 2.52, 2.93 and 3.06 for grana, grana margin and Stroma lamellae, respectively.

Why the ratio was very low for Grana margin and Stroma Lamellae? At which step I did wrong or what could be wrong?

Method: I extracted the crude thylakoid directly by a method as described in the Surosa et al. (2006), JBC; 281;145-150. and then solubilized the membrane according to the method used by Bhuiyan et al. 2015; Plant Cell, 27; 262-285. Membrane (0.5 mg/ml Chl dilution) in buffer (25 mM trichinae pH 7.8, 150 mM sorbitol, 10 mM NaCl and 5 mM MgCl2) were solubilized by 1% digitonin at final concentration, for 15 min on ice. After removing the unsolubilized fractions by spinning at 1000 g for 3 min, grana core was pelleted at 10000 g for 30 min, grana margin was pelleted at 40000 g for 30 min and finally Stroma lamellae was pelleted at 145000 g for 60 min.