Geri Dön

Verification of ingredient labels in high-risk oils and fruit juices by using vibrationalspectroscopy combined with pattern recognition analysis

Yüksek riskli yağlar ve meyve sularının içerik etiketlerinin titreşim spektroskopisi ve örgü tanıma analizi ile doğrulanması

PDF İndir

  1. Tez No: 784593
  2. Yazar: DİDEM PEREN AYKAS
  3. Danışmanlar: PROF. DR. LUIS RODRIGUEZ-SAONA
  4. Tez Türü: Doktora
  5. Konular: Gıda Mühendisliği, Food Engineering
  6. Anahtar Kelimeler: Belirtilmemiş.
  7. Yıl: 2019
  8. Dil: İngilizce
  9. Üniversite: The Ohio State University
  10. Enstitü: Yurtdışı Enstitü
  11. Ana Bilim Dalı: Belirtilmemiş.
  12. Bilim Dalı: Belirtilmemiş.
  13. Sayfa Sayısı: 184

Özet

Özet yok.

Özet (Çeviri)

Food adulteration and counterfeiting is a major worldwide problem with a cost of as much as $15 billion annually and affecting nearly 10% of all food products on the market. The consequences of the economic adulteration including reduced market size, increased costs related to a recall, liability, withdrawals, lost revenue or market share, damaged brand, and even sometimes failed business or bankruptcy. Besides its economic impact, public health risks could cause far more consequences to the related food industry or food company. Food fraud has been conducted since ancient times, and it is still a worldwide public concern, and a leading cause of trade problems internationally, olive oil and wine were the first counterfeit foods followed by fruit juices, spices, tea, milk, honey, and saffron. Counterfeiters target high-value products and products with a strong brand name, because of a more substantial profit possibility, resulting in potentially dangerous counterfeit with ingredients unlikely to have been subjected to the laborious quality control for food products. Food fraud undermines the economy but also may cause the potential public health risks. In 2007, melamine-contaminated pet foods killed dogs and cats in the United States, and after that incident, melamine-contaminated infant formula and milk powder hospitalized hundreds of thousands and killed six babies in China. In 1981, olive oil iv adulteration with colza oil that contains aniline derivatives killed over 600 people and affected over 20,000 people in Spain. United States Food & Drug Administration (FDA), United States Department of Agriculture (USDA) and other international governmental agencies (i.e., European Commission, Codex Alimentarius) have been working to protect the food supply from food safety risks (both intentional or unintentional contamination). However, highly time-consuming and labor-intensive traditional techniques hamper these efforts. As an alternative technology, vibrational spectroscopy offers simple, fast, specific and sensitive analysis along with minimal or no sample preparation. Advances in vibrational spectroscopy instruments including Fourier transform infrared (FT-IR) and Raman spectroscopy combined with pattern recognition analysis techniques have made possible rapid material screening (~1 min) with minimal sample preparation and personnel training. Fingerprinting approaches for untargeted detection of economic adulteration can offer the food industry the ability to screen the presence of contaminants providing the food industry unique detection capabilities to prevent fraud and ensure consumer product safety. Implementation by the industry and regulatory agencies of rapid testing procedures based on these technologies would help to streamline quality assurance, and food safety is preventing the growing danger to the health of consumers from adulterated or substituted products as evidenced by the melamine incident. The overall objective of this study was to establish a reliable ingredient label verification program(s) for edible oils and fruit juices using portable mid-infrared and Raman spectroscopic techniques combined with pattern recognition analysis. We v employed untargeted and targeted approaches for label verification depending on the specific aim of the work. Our first aim evaluated an untargeted approach for authentication of edible oils used in the manufacturing of potato chips by combining the fingerprinting capabilities of a portable 5-reflection attenuated total reflectance infrared (FT-ATR) spectrometer combined with supervised pattern recognition. We collected 105 samples of potato chips from local markers and online suppliers that encompassed a wide variety of frying oils either indicating the sole source or the use of one or more oils. Oils were characterized by reference methods that included fatty acid composition (GC-FAME), free fatty acids (FFA), peroxide value (PV) and p-anisidine values. Based on the fatty acid composition, we identified ten different frying oils that were used by manufacturers in producing the potato chips including (corn, mid-oleic (MO) sunflower, high oleic (HO) sunflower, expeller-pressed sunflower, HO safflower, canola, two types of HO canola, cottonseed, peanut). A prediction algorithm was developed using soft independent of class analogy (SIMCA) that clustered the samples (n=69) based on a single source of oil. A validation set (n=13) using samples that were not included in the training set showed excellent sensitivity and specificity of the models with 100% accurate predictions and no miss-classifications based on the fatty acid composition data. Furthermore, the model correctly indicated samples manufactured with a mixture of oils as observations that did not belong to any of the classes. Also, our algorithm identified three samples were mislabeled of the oil source. Furthermore, the same spectra that collected from the oils were used to develop partial least square regression models (PLSR) to predict oil quality parameters, and strong correlations between the reference test results and infrared predicted values were found. The second study aimed to develop a non-targeted approach to authenticate EVOOs using vibrational spectroscopy (FT-IR and Raman) in combination with pattern recognition analysis. For this purpose, a total of 151 commercial samples that were labeled as“EVOO”were obtained from 9 different countries (Turkey, California, Italy, Spain, Greece, Turkey, Tunisia, Portugal, and Peru). The samples were classified in 4 different groups according to their information obtained from Olive Oil Research Laboratory (Aydin, Turkey), and California Olive Oil Council, also the reference data that we collected. Olive oil samples were grouped as EVOO (n=77), virgin olive oil (VOO) (n=10), lower quality olive oils (blends of refined and virgin olive oils) (n=17), and olive oils adulterated with vegetable oils and olive pomace (n=47). The spectra were collected using a portable five-reflections FT-IR and a Raman spectroscopy with 1064 nm excitation laser. The SIMCA models gave best classification performance for EVOO by using FT-IR spectra showing high discrimination from VOO (ICD: 3.2), lower quality olive oils (ICD: 2.7), and the adulterated olive oils (ICD: 5.2). On the other hand, the SIMCA model that was generated using the Raman spectra had lower sensitivity on the samples with similar profiles (virgin and refined olive oils) but allowed detection of adulteration with vegetable oils (ICD: 3.2). The models were validated using an external validation set (n=30), and EVOO samples were correctly identified with a 100% accuracy. Furthermore, a targeted approach was used to predict olive oil key quality characteristics (fatty acid composition, free fatty acids, peroxide value, total polar vii compounds, and pyropheophytins) by combining the collected spectral information and reference data. Regression models showed excellent correlation (Rval0.96) and low standard error of cross-validation (SECV) both with FT-IR and Raman spectroscopy. In the last study, the aim was to develop a targeted prediction models for determination of multiple quality traits (sucrose, glucose, fructose, and total sugars, ascorbic and citric acids, titratable acidity, and soluble solids) of fruit juices (FJs) by using a field-deployable and portable FT-IR spectroscopy with no sample preparation. A total of 68 fruit juice samples with a single (n=25) or multiple (n=43) fruit origins were collected from US markets. The quality traits of the samples were determined using official analytical methods, and FT-IR spectra were collected using a portable FT-IR with a transmission accessory. In FJs, the primary calorie intake is from the sugar content, and an excessive amount of sugar has a correlation with obesity and over-weight. On the other hand, fruits and vegetables supply about 90 percent daily ascorbic acid, which has a wide range of beneficial health effects and essential for the collagen synthesis in the body. Some packaged foods including fruit juices are promoting through health and nutrition claims that states“ascorbic acid promote to support a healthy immune system,”and the health and nutrition claims are one of the most important attributes to a consumer choice. Therefore, for a fruit juice, it is essential to have a nutrition label that complies with the levels that were reported by the FJ producers. Soluble solids (ᵒBrix), titratable acidity, citric acid and ᵒBrix/acid ratio are the main physicochemical characteristics, besides the sugar and ascorbic acid content of a juice in order to determine the quality. The PLSR models were developed to predict the quality traits by combining the FT-IR spectra and the calculated reference data. Overall, the PLSR models showed good correlation (Rpred≥0.94) and low SECV between the predicted and the measured values. We also compare the declared value on the nutrition labels and the reference results, and it had been found that 15% and 40% of FJs were not in compliance with the declaration for total sugars and ascorbic acid, respectively.

Benzer Tezler

  1. Tez No

    134281

    Use of multivariate statistical techniques in HACCP programs

    Çok değişkenli istatistiksel tekniklerin HACCP programlarında kullanılması

    UMUT BAŞAK BALIKLI

    Yüksek Lisans

    İngilizce

    İngilizce

    2003

    Gıda Mühendisliğiİzmir Yüksek Teknoloji Enstitüsü

    Gıda Mühendisliği Ana Bilim Dalı

    YRD. DOÇ. DR. FİGEN TOKATLI KÖSEBALABAN

  2. Tez No

    202122

    Balık yemlerinin ve yem hammaddelerinin kalitesinin mikroskobik yolla incelenmesi

    Microscopy of fish feed and fish ingredients quality

    DENİZ ERGEN

    Doktora

    Türkçe

    Türkçe

    2007

    Su ÜrünleriEge Üniversitesi

    Su Ürünleri Yetiştiriciliği Ana Bilim Dalı

    YRD. DOÇ. DR. AYSUN FIRAT KOP

  3. Tez No

    467211

    Yaşlıların düşme tespiti ve bilgilendirme sistemi tasarımı

    Fall detection and notification system design for elderly people

    FİKRİ ELMAS

    Yüksek Lisans

    Türkçe

    Türkçe

    2017

    Biyomühendislikİstanbul Teknik Üniversitesi

    Elektronik ve Haberleşme Mühendisliği Ana Bilim Dalı

    DR. SERKAN TÜRKELİ

  4. Tez No

    783511

    İkili kuadratik form ̇ile grup kimlik doğrulaması

    Group authentication with binary quadratic form

    FİLİZ AKSOY

    Yüksek Lisans

    Türkçe

    Türkçe

    2023

    Bilim ve Teknolojiİstanbul Teknik Üniversitesi

    Bilişim Uygulamaları Ana Bilim Dalı

    PROF. DR. ENVER ÖZDEMİR

    DOÇ. DR. ÖZEN ÖZER

  5. Tez No

    911647

    Aronya (Aronia melanocarpa) meyvesinden değerli biyoaktif bileşenlerin ekstraksiyonu ve ağızda eriyen ince şerit üretiminde kullanılması

    Extraction of valuable bioactive compounds from black chokeberry (Aronia melanocarpa) fruit and using in production of oral dissolving strips

    CEREN EVRİM MATARACI

    Yüksek Lisans

    Türkçe

    Türkçe

    2024

    Gıda MühendisliğiEge Üniversitesi

    Gıda Mühendisliği Ana Bilim Dalı

    DOÇ. DR. SEHER KUMCUOĞLU

Geri Dön