CircularDichroism例析.ppt

Bandwidth (SBW) selection Setting of slits should be as large as possible (to decrease noise level), but compatible to the natural bandwidth (NBW) of the bands to be scanned. As a rule SBW should be kept at least 1/10 of the NBW, otherwise the band will be distorted. If NBW is not known a series of fast survey spectra at different SBW will help proper selection. Trade in of accuracy versus sensitivity (i.e. the use of larger than theoretical SBW) is occasionally required. 2 nm in the far UV region 1 nm in the aromatic region (where fine structures may be present), optimal band-pass (as large as possible, but not loosing information) can be determined after a trial Number of data point data pitch, i.e. number of data points per nm, will not directly influence the noise level. However if post run further data processing will be applied to reduce the noise, it’s advisable to collect as many data points as possible to increase the efficiency of the post run filtering algorithm Accumulation another way to improve S/N is to average more spectra. Here too the S/N will improve with the square root of the number of accumulations. Averaging is very effective since it compensates short term random noise, but it’ll not compensate long term drifts (mainly of thermal origin). So if long accumulations are used we recommend a suitable long warm-up of the system and/or the use of a sample alternator (to collect sequentially sample and blank and average their subtracted values). For long overnight accumulations it’s essential that room temperature is well kept stable. Sample concentration and cell pathlength A good suggestion is to run in advance an absorption UV-VIS spectra. CD spectroscopy calls for same requirements as UV-VIS: best S/N is obtained with absorbance level in the range 0.6 to 1.2. It’s usually difficult to get proper data when absorbance (of sample + solvent) is over 2 O.D. Typical Conditions for protein CD Protein Concentration: 0.2 mg/ml Cell Path Length: 1 mm Vol