The Quality Evaluation Of Saffron By Near-infrared Spectroscopy

Saffron,the dried stigma of Crocus sativus L.primarily distributes in areas like Iran,Italy,Spain and Greece.As an expensive spice and a traditional herb,saffron has drawn considerable attention to its quality.Picrocrocin,safranal and crocins are the main chemical foundation of the bitterness,scent and color of saffron respectively.And they are always used as indices to estimate the quality of saffron.UV-VIS,HPLC and GC were the main analytical methods for the quality evaluation of saffron.Near-infrared spectroscopy(NIRS)is a rapid,non-destructive and informative analysis technique which has been applied widely in the quality evaluation of Traditionanl Chinese Medicine(TCM),recently.For example,the NIRS quantitative and qualitative models can be used to predict the contents of chemical compositions and discriminate the difference of geographies and categories of TCM.Thus,the establishment of NIRS method which is used to estimate the quality of saffron will be a meaningful work for the quality evaluation system of saffron.In the present study,NIRS together with conventional analysis methods were implemented to estimate the quality of the 98 batches of saffron samples which were purchased or harvested from Iran and China.And the main contents of the research are as followings:1.Review of the research progress of saffron.The correlative literatures have been presented and analyzed,including the resources,chemical components,quality evaluation and pharmacological effects,etc.2.The commercial classification of saffron by ISO method.The determination of moisture and volatile matter and the identification of morphological character were carried out according to Chinese Pharmacopoeia(2015)methods.The spectrophotometric method specified in ISO 3632-2:2010 was used to determine the bitterness,scent and color of saffron,and the three values were used to judge the commercial categories of samples.It was found that the batches of category?,category?,category ? and unqualified samples were 24,36,30 and 8,respectively.Meanwhile,colorimetry(L*,a*,b*)technique was used to analyze the colour information of saffron.A good relationship between a* and the absorbance of crocins / the color of sample was obtained in the research.3.The qualitative and quantitative analysis of chemical compounds of saffron by HPLC-DAD/MS.Eight common peaks were exhibited in HPLC fingerprint of 98 batches of saffron,and six of them were identified,namely picrocrocin,trans-crocin 4(TC 4),trans-crocin 3(TC 3),cis-crocin 4,cis-crocin 3 and safranal,respectively.The poor similarity of HPLC fingerprint obtained at 308 nm showed that there was some difference among these samples.Multivariate analysis was used to explore the difference between saffron from China and Iran.Chinese saffron was turned out to be richer in the contents of safranal,TC 3 and picrocrocin,while poorer in the contents of cis-crocin 3(P<0.05).And the content of TC 4 and cis-crocin 4 showed no difference between the samples from China and Iran(P > 0.05).Additionally,an HPLC method was established to simultaneously determine the three major ingredients,namely TC 4,TC 3 and safranal.The results showed that the contents ranges of TC 4,TC 3 and safranal were 4.24~17.42%,1.84~7.92% and 0.06~0.57%,respectively.4.The quality evaluation of saffron via NIRS.In the research,NIRS quantitative model for the content of moisture and volatile matter,NIRS quantitative models for the contents of index components and the geographical discrimination NIRS qualitative model were established to estimate the quality of saffron.The results showed that the prediction performance of moisture and volatile matter NIRS quantitative model was excellent(the values of RMSEP,RMSECV and RPD were 0.228,0.271 and 8.86,respectively);the prediction performance of TC 4,crocins(the sum of TC 4 and TC 3)and picrocrocin(peak area/unit mass)NIRS quantitative models were all satisfactory(the RMSEP values of the three models were 0.770,0.905 and 0.696;the RMSECV values of the three models were 0.758,0.890 and 0.695;the RPD values of the three models were 4.14,4.32 and 4.48);the prediction performance of TC 3 and safranal NIRS quantitative models were average(the RMSEP values of the two models were 0.366 and 0.0355;the RMSECV values of the two models were 0.410 and 0.0408;the RPD values of the two models were 2.97 and 2.54,respectively).The two models can be used to predict the contents of TC 3 and safranal,but the accuracy of prediction need to be improved.For the geographical discrimination NIRS qualitative model,the positive rates of calibration and prediction samples were 97.22% and 96.15% which indicated that the NIRS qualitative model can be used to discriminate the saffron from Iran and China.Above all,crocins(TC 4 and TC 3),safranal and picrocrocin were used as index components of saffron in the research.Meanwhile,UV-VIS,colorimetry(L*,a*,b*)analysis,HPLC and HPLC fingerprint methods,NIRS technology were used to evaluate the quality of saffron,such as commercial category,the qualitative and quantitative analysis methods of chemical compounds and the establishment of NIRS models.