核磁能检测含量吗? 致汪泉：提供下面短文，供你参考。 instructions for quantitative nmr spectroscopy quantitative 1h spectra in order to obtain accurate integrals of one-dimensional 1h spectra, the following parameters must be set: relaxation delay - this is the delay between successive transients and must be long enough to allow the magnetization to achieve equilibrium along the z-axis (approximately 5* t1 for the proton with the longest t1). spectral width (sw) and consequently the filter bandwidth (fb) must be sufficiently large so that peaks at the edges of the spectrum are not attenuated by the filter. to avoid this problem, set the sw so that there is at least 2.5 ppm between your peaks and the edge of the spectrum and open up the filter bandwidth by setting fb=sw. digital resolution - peak integrals involve the summation of the points defining your spectrum, therefore, for the most accurate summations your digital resolution should be about 10 points/hz. the parameter np specifies the number of complex points collected, therefore for sufficient digital resolution use np = 10*sw and zero fill the final spectrum once by setting fn=2*np. the width of the pulse should be a little less than 90 so set pw=7 at tpwr=58. for the af-300 set the flip parameter to 7 u. on varian spectrometers apply a linear baseline (drift) correction (type dc) before working up the integrals. (this correction is automatically applied on the af-300 with the (apple) k command). this will flatten your baseline as long as you don't have peaks too close to either ends of the spectrum. it will automatically set the lvl and tlt parameters which are the zero- and first-order parameters for baseline correction. you can also change these manually by clicking on the lvl/tlt button in the ds or display interactive menu. for integrals of peaks, you must select the regions of the spectrum over which the points will be summed. first click on part integral in the ds menu. if you have peaks separated by well defined areas of baseline type region for automatic selection of integral points. if your spectrum is more crowded you will have to select integral regions manually. to do this type cz to clear current integral resets then click on resets. now specify the regions over which to integrate by clicking once on the baseline to the left of a peak and then on the baseline to the right of the peak. the integral line should be solid over the peak and dotted where there is baseline. perform this function for all peaks. if further baseline correction is required and there are no peaks that haven't been selected for integration type bc for non-linear baseline correction. depending on the peaks selected this command can cause severe base-line distortion. when this occurs you're better off using only the linear baseline correction command dc. type wft to retransform the spectrum and then type dc. to display integrals on screen: set vp=15 type dpirn to display normalized integral regions below spectrum. for normalized integrals the sum of all values is equal to the value of the ins parameter. you can change this value by typing ins=#. type dli to display a list of integrals to assign a value to a particular integral use the command setint(int#,value). now use dpir to display integrals underneath peaks or use dli. to print integral values to plot spectrum type pl pscale to plot integral values below spectrum type pirn(normalized integrals) or pir if you used the setint command. type page to advance the paper from printer. quantitative 13c spectra integration of 13c spectra is more complicated than integrating 1h spectra. 13c relaxation times tend to be very long so your spectrum may require using a very long relaxation delay time (d1 or last delay). these relaxation times may also be decreased by adding a relaxation agent such as cr(acac)3. noes between 1hs and 13c nuclei will also affect your measured intensities so you must use inverse-gated decoupling which involves decoupling 1hs only during acquisition. for inverse-gated decoupling type dm='nny' on the varian instruments. on the af-300 open up the pulse sequence icon(click on far right icon in upper left-hand corner). open the file called 1 pulse dec on acq. for the most accurate integrals you want to make sure the relaxation delay is very long relative to the longest t1 in your sample. this delay should be at least 2.5xt1 and at least 10 times longer than the acquisition time. the s/n ratio should be at least 35:1 for the best results.查看更多
第1天
关注盖德视界
添加小助手
