Investigations on the Chemistry of Coffee Processing Seung-Hun Lee Doctor of Philosophy in Chemistry

Home , Black Ivory coffee, Coffee filter, Kahweol, Korean tea ceremony

Investigations on the Chemistry of Coffee Processing

Seung-Hun Lee

a Thesis submitted in partial fulfillment of the requirements for the degree of

Doctor of Philosophy in Chemistry

Approved Dissertation Committee

Prof. Dr. Nikolai Kuhnert Name and title of Chair Prof. Dr. Thomas Nugent Name and title of Committee Member Prof. Dr. Adam Le Gresley Name and title of Committee Member

Date of Defense: 03. 09. 2018

Statutory Declaration

Family Name, Given/First Name Lee, Seung-Hun

Matriculation number 20330669 What kind of thesis are you submitting: PhD Thesis Bachelor-, Master- or PhD-Thesis

English: Declaration of Authorship

I hereby declare that the thesis submitted was created and written solely by myself without any external support. Any sources, direct or indirect, are marked as such. I am aware of the fact that the contents of the thesis in digital form may be revised with regard to usage of unauthorized aid as well as whether the whole or parts of it may be identified as plagiarism. I do agree my work to be entered into a database for it to be compared with existing sources, where it will remain in order to enable further comparisons with future theses. This does not grant any rights of reproduction and usage, however.

This document was neither presented to any other examination board nor has it been published.

German: Erklärung der Autorenschaft (Urheberschaft)

Ich erkläre hiermit, dass die vorliegende Arbeit ohne fremde Hilfe ausschließlich von mir erstellt und geschrieben worden ist. Jedwede verwendeten Quellen, direkter oder indirekter Art, sind als solche kenntlich gemacht worden. Mir ist die Tatsache bewusst, dass der Inhalt der Thesis in digitaler Form geprüft werden kann im Hinblick darauf, ob es sich ganz oder in Teilen um ein Plagiat handelt. Ich bin damit einverstanden, dass meine Arbeit in einer Datenbank eingegeben werden kann, um mit bereits bestehenden Quellen verglichen zu werden und dort auch verbleibt, um mit zukünftigen Arbeiten verglichen werden zu können. Dies berechtigt jedoch nicht zur Verwendung oder Vervielfältigung.

Diese Arbeit wurde noch keiner anderen Prüfungsbehörde vorgelegt noch wurde sie bisher veröffentlicht.

………………………………………………………………………………………………………. Date, Signature

g{x g{xá|á

Wxw|vtàxw àÉ lÉâÇz@@[xx[xx ctÜ~

TÇw

]tx@@[çâv~[çâv~ _xx

Table of Contents Table of Contents ...... 4 Abstract ...... 6 Acknowledgement ...... 8 Abbreviations ...... 11 Achievements ...... 12 Curriculum Vitae ...... 14 Preface ...... 20 1. Introduction ...... 21 1.1. Coffee Chemistry ...... 21 1.1.1. Coffee Compounds ...... 21 1.1.2. Chlorogenic acids ...... 22 1.1.2.1. Etymology and classification ...... 22 1.1.2.2. Polyphenols ...... 23 1.1.2.3. Chlorogenic acids in coffee ...... 24 1.1.2.4. Analysis of chlorogenic acid ...... 36 1.1.2.5. Properties of chlorogenic acid ...... 38 1.1.3. Caffeine ...... 39 1.1.3.1. Etymology and classification ...... 39 1.1.3.2. The role of caffeine ...... 40 1.1.3.3. Coffee decaffeination ...... 41 1.2. Coffee Technology ...... 42 1.2.1. Coffee Cherry ...... 42 1.2.1.1. Classifications ...... 42 1.2.1.2. Processes ...... 43 1.2.2. Green Coffee ...... 44 1.2.2.1. Classifications ...... 44 1.2.2.2. Processes ...... 48 1.2.3. Roasted Coffee ...... 50 1.2.3.1. Classifications ...... 51 1.2.3.2. Processes ...... 53 1.2.4. Ground Coffee ...... 54 1.2.4.1. Classifications ...... 54 1.2.4.2. Processes (Brewing / Extraction) ...... 55 1.2.5. Extracted Coffee ...... 55 1.3. Coffee Sensory Perception ...... 56 4

1.3.1. Coffee Aroma ...... 56 1.3.2. Coffee Taste ...... 58 1.3.3. Tools for Coffee Sensory Analysis ...... 59 1.3.3.1. World Coffee Research Sensory Lexicon ...... 59 1.3.3.2. The Coffee Taster’s Flavor Wheel ...... 60 2. Aim of the Research ...... 70 3. Results ...... 72 3.1. Research on Green Coffee Step ...... 72 3.2. Research on Roasting Step ...... 89 3.3. Research on Brewing Step ...... 125 3.3.1. Research on Brewing Step with Different Solvents ...... 125 3.3.2. Research on Brewing Step with Different Methods ...... 139 3.4. Research on Application to Other Food ...... 167 4. Conclusion ...... 194

Abstract

The International Coffee Organization announced that coffee consumption has increased for 4 years consecutively by 1.9 % every year from 2013/14 to 2016/17. The economic importance of coffee is continuously increasing, so that research on coffee becomes more and more important. Chlorogenic acids (CGAs) and caffeine are the best investigated components among all the coffee compounds after Clifford group developed easy identification method by using novel LC-MS n at 2003. They have been identified as the most health-affecting compounds in coffee.

A lot of coffee research has investigated the chemical composition of green or roasted coffee beans and the final cups of coffee by academics and industries, respectively. However, still there is little research studying the role of composition changes caused by each individual coffee processes. Furthermore, increasing demands of specialty coffee promotes new developments of coffee processes but the inventors need scientific input for analyzing the component with bioactivities to support and market their innovations.

In this thesis, the different stages of coffee processing have been investigated using suitable experiments in those stages; cultivating, grading, roasting and brewing steps were investigated. Additionally, those methods were applied to analyze Korean traditional green tea samples.

The different grades of Jamaican Blue Mountain coffee, a product of high price and rarity, were analyzed by LC-MS n together with different originated common coffees. The relative ratios of regioisomers of caffeoylquinic acids (CQAs) were determined as the possible specialty flavor for the first time from this source.

A new coffee product marketed as fermented coffee, obtained by a consecutive steaming and drying process was also investigated. This product, which is marketed as a roasting substitute for coffee, showed the highest antioxidant property if compared to the green coffee bean and two roasted coffee beans with different degree of roasting. This was the first time reported that the heating processed coffee has higher antioxidant property than the green coffee bean has. The differences of chemical composition of this new process were also identified by using LC-MS n method.

For the brewing step of coffee process experiment, the different composition ratios of chlorogenic acids by changing extracting solvent, which is easily predicted but not yet scientifically established and published, were identified. Moreover, cold-brew coffee, which is a recent coffee brewing method, is analyzed for the first time in this study.

By combining all the coffee experiments, Korean green tea samples were analyzed by the same LC-MS n method as used in coffee analysis. The high-priced special tea samples were selected as Jamaica Blue Mountain coffee is in coffee and the producing method includes consecutive heating and drying process as fermented coffee in coffee analysis. Additionally, different extracting solvents were used as brewing step experiment in coffee analysis. To sum up, all the key-points in the coffee experiments in this thesis were applied to other food analysis and could find the applicability of the methods as components controller factors.

Through these studies, regioisomeric ratio of chlorogenic acids were suggested as a quality influencing factor, a coffee processing method with higher antioxidant property was found, the extraction influences on the kind and quantity of solvents were identified and the applicability of all the methods from coffee to other food techniques was confirmed.

Acknowledgement

First of all, I would like to begin by expressing my gratitude to Prof. Dr. Nikolai Kuhnert, not only as my supervisor of my Ph.D. research, but also my advisor of my life in Germany for work/life balance, linguistic development and being a father in a family. Through his guidance, encouragement and motivation, I could develop myself as an analytic scientist, although it is totally different field I had done in Korea as a synthetic chemist. Moreover, I could settle in Germany as a leader of 3-membered family.

I would also like to acknowledge Prof. Dr. Thomas Nugent and Prof. Dr. Adam Le Gresley for taking time to review my thesis. Prof. Nugent made me maintain as a scientist when I lean toward coffee field during entire of my Ph.D. period. And Prof. Adam Le Gresley gave me the insight, being as an independent scientist from Prof. Kuhnert’s guidance.

I would like to thank to sending me the coffee and tea samples from all over the world for my research. The people working in the Embassy of the Republic of Korea in Jamaica and Young-gi Moon helped me Jamaica Coffee Research by sending regional samples. Cafe Romeo helped me Cold-brew Coffee Research by sending their brewing machines. Coffeebio helped me Fermented Coffee Research by sending their Fermented coffee samples and also helped me to compare with different roasted coffee samples. Geumro helped me Korean Traditional Green Tea Research by sending their consecutive parched green tea samples. All the supports did not charged me anything, genuinely supported for the improvement of scientific food understanding.

Furthermore, Genesis and Lloyd coffee, though it was not free, provided me a special discount and fruitful informations of the products for my research. Genesis provided their 100 ~ 250 g coffee roaster, which is perfect for the lab-scale research. Lloyd coffee provided a variety of green coffee bean samples in different origins. Additionally, the people working in the Embassy of the Republic of Korea in Yemen, Eunsang Kim, Coffee and Tea, Seung-Kuk Park, Nahee Kim, Joon-Kwan Moon, Juho Choi, Stronghold, Jung-Im You, Byunggun Park and Jikwang Han did not work together with me, but gave fruitful advices for my coffee research.

And there were a variety of associations which I would like to express my gratitude to. Starting from Jacobs University Bremen for providing me a nice environment for my research, DAAD for scholarship, GDCh for scholarship to conferences, VeKNI for making me feeling 8

at home in Germany and scholarship, Cocotea conference for providing continuous opportunities to keep me up-to-date on coffee research and Kids at Jacobs for letting me concentrate on my research out of the babycare.

I would like to individually thank all of my colleagues. Starting with our lab’s excellent technicians, Anja Mueller for her total-care for our lab core facilities, should be firstly referred because all our lab’s research could not have done without her contribution. Britta Behrends established our lab’s systematic management system, so directly or indirectly, all our lab’s results are supported by her contribution. And I thank to those people worked together, Dr. Rakesh Jaiswal for the instructions of MS, Dr. Roy Dsouza for physical - calculational help, Dr. Sagar Deshpande for his Energy to our lab, Dr. Hande Karaköse for her lab tea-time management, Dr. Agnieszka Golon for her happiness to the lab, Dr. Mohamed Elsadig for his warm-invitation and interesting episodes, Dr. Matei, Marius-Febi for his interesting stories for German football system, Dr. Maria Alexandra Patras for her exemplary attitudes and a lot of practical advices of baby care, Dr. Abhinandan Shrestha for taking the most time in the office without drinking coffee, Dr. Rohan Shah for his encouraging me to brew better coffee and pleassure talks about coffee, Dr. Inamullah Hakeem Said for his contributions for managing lab members, Dr. Diana Sirbu for showing diligent attitudes to make me motivated, Dr. Roberto Megias for his friendly discussions and the honey-tea gift from Spain, Anastasiia Shevchuk for her contributions for lab-teatime and helps for visiting Asian Cocotea, Pawel Andruszkiewicz for being the center for connecting before and after students, Maria Bikaki for her contributions of members’ birthdays, Sabur Badmos for being my direct theme-successor and made a lot of fruitful results to fulfill my curiosity on coffee, Fariba Sabzi for interesting informations of nuts and showing improvement of her lifestyle, Tahira Musarrat for balancing our lab’s theme with her background of synthesis, Hossameldeen Sharaf for making me comfortable even though we did not have enough time together. Also, I would like to thank to Euisuk Han and Suhyun Moon, for being my students to work together with low acidic coffee and human Luwak coffee projects. Additionally, these people made my Ph.D. period happier even though we did not work together; Beatrix Tetzlaff, Prof. Hartwig, Dr. Mubo, Prof. Lalith, Borislav Milev, Dr. Ghada Yassin, Nickolet Ncube, Dr. Jonathan Cave and Dr. Maria Isbel Alarcon.

I would like to thank the Korean communities in Bremen to make me feel at home with Korean language environment. I thank to Jacobs Korean community to make regular gathering twice per year, to make Korean foods and good companies. Starting from the big

brother Jonghak who has drunk coffee the most times together with me, Aemin, Seo-bros, Hyeyoung, Eunhye, Hanbit, Hojun who kept me as an early adapter, Minseok, Sojung who was a nice consultant with drinking coffee, Haeun, Ikyung, Suhyun, Euisuk, Myunggi, Dain, Jangsup, Jaeho, Youngjun, Hyunseok and all the Koreans I could not mention here made my Ph.D. period more comfortable. Also, it was a great consolation that the two Korean churches and Bremen Korean society were always in the backup whenever the Koreans needed any types of helps. Especially the times with the coincidence-inevitability friend Hyoseup and Hyejin, Jung’s family who helped a lot for my starting period in Bremen, Sujin and Wang’s family, Sanga and Hyunggu are my indelible memories. Also, the three Korean restaurants in Bremen were very helpful whenever I wanted to taste foods from Korea.

The friends, always be there and make me happy from Korea, are also appreciated. My formal colleges of Samyang Group, Kwangsu, Kyungdo, Wonjun, Sanghun and the friends of my childhood, Sangjin, Myunggyu, Kyunghwa and the friends of my University periods, Suyeon, Hyejin and the friends of my English studies were always there, made me feel at home. Especially, Kyunghwa and Hyejin visited Germany during my Ph.D. periods, so I want to express my thanks again here. And I would like to thank to Kyung-ju Hwang for his advices for the health of myself and my families.

I would like to thank to Prof. Jung-Il Jin, who was the advisor of my master degree, since he always gives me advices for my future, and supports me as much as he can. Together with him, the members of Gogohoi would also be appreciated with their warm words and cares from Korea.

Last but not the least, I thank to my family (and family-in-law) in Korea and the family living in Berlin and Essen in Germany. I thank to especially my parents, sister, Susie, Woosung, and Junghun for physically coming from Korea to Germany to see us. Although the other family members could not visit us, I know that they always care for us, and welcome me whenever I visit Korea from time to time. I thank to Minjee and Subin to be always with me, be my side and will always be in the future.

Abbreviations

BPC Base Peak Chromatrogram CA Caffeic Acid CCoQA Caffeoylcoumaroylquinic Acid CFQA Caffeoylferuloylquinic Acid CGA Chlorogenic Acid CGAs Chlorogenic Acids CoQA Coumaroylquinic Acid CQA Caffeoylquinic Acid CQL Caffeoylquinic Acid Lactone CSA Caffeoylshikimic Acid diCQA Dicaffeoylquinic Acid DFQA Dimethoxycinnamoylferuloylquinic Acid DQA Dimethoxycinnamoylquinic Acid EIC Extracted Ion Chromatrogram ESI Electrospray Ionisation FQA Feruloylquinic Acid FSQA Feruloylsinapoylquinic Acid FTMQA Feruloyltrimethoxycinnamoylquinic Acid HPLC High Performance Liquid Chromatography M-CQA Caffeoylquinic Acid Methyl Ester M-diCQA DiCaffeoylquinic Acid Methyl Ester MeOH Methanol MS Mass Spectrometry NMR Nuclear Magnetic Resonance QA Quinic Acid SiQA Sinapoylquinic Acid TIC Total Ion Chromatogram TOF Time of Flight

Achievements

Publications

• Seung-Hun Lee, Rakesh Jaiswal, Nikolai Kuhnert, Analysis on different grades of highly-rated Jamaica Blue Mountain coffees compared to easily available originated coffee beans, SCIREA Journal of Food. Vol. 1 , No. 2 , 2016 , pp. 15 – 27 • Seung-Hun Lee, In Jae Kim, Sabur Badmos and Nikolai Kuhnert, Analysis of Steamed (“Fermented ”) Coffee Beans as a Roasted Coffee Substitute, 2018, in preparation • Seung-Hun Lee, Nikolai Kuhnert, Comparision of the chlorogenic acids content of laboratorial and practical solvent extracted coffees, 2018, in preparation • Seung-Hun Lee, Nikolai Kuhnert, Brewing Less- and Enhanced-caffeinated Coffee by Using Cold Brew Method, 2018, in preparation • Seung-Hun Lee, Inamullah Hakeem Said, Un-Jae Lee, Nikolai Kuhnert, Analysis on kinetic parching samples of Korean traditional green tea process with different brewing methods, 2018, in preparation

Poster Presentations

• Characterization of Three Different Grades of Jamaica Blue Mountain Coffee by LC- MS n Second International Congress on Cocoa Coffee and Tea, 2013, Naples, Italy • Analysis of Fermented Coffee Beans as a Roasted Coffee Substitute Bremen Life Sciences Meeting 2015, 2015, Bremen, Germany • Identification of Composition Kinetics from Cold-brewed Coffee by LC-MS n Third International Congress on Cocoa Coffee and Tea, 2015, Aveiro, Portugal • Identification of Composition Kinetics of Traditional Korean Green Tea Producing Method by LC-MS The Korean Society of Food and Nutrition Summer Conference, 2016, Seoul, Korea • Analysis of Less- and Enhanced- Caffeinated Coffee: LeeN-Caffeination Korea Society of Food and Cookery Science, 2017, Seoul, Korea • Identification of composition kinetics on applications of the LeeN caffeination coffee 12

Fourth International Congress on Cocoa Coffee and Tea, 2017, Turin, Italy

Articles in Popular Press

• Compounds to diabetes in coffee and tea, Coffee & Tea , 2017. 02. • Polyphenols and antioxidant effect, Coffee & Tea , 2017. 04. • Flavor of coffee? Minimization of Furans?, Coffee & Tea , 2017. 06. • Is really Dutch coffee decaffeinated? Are more than 10 % coffee companies unscrupulous?, Coffee & Tea , 2017. 08. • Is coffee good for health or not? Why does the debate still go on?, Coffee & Tea , 2017. 10. • The great scholars on coffee research, Coffee & Tea , 2017. 12. • Coffee addiction? Caffeine addiction? The story of addiction and dopamine, Coffee & Tea , 2018. 02.

• Trigonelline and Vitamin B 3 in Coffee, Coffee & Tea , 2018. 04., submitted.

Books

• Chlorogenic acids in Coffee, Book chapter invited, The Royal Society of Chemistry (RSC)

Awards

• 1st Prize , Home Brewers Cup Bremen, Bremen, 2014 • Excellent Poster Prize , The Korean Society of Food and Nutrition, Seoul, 2016

Curriculum Vitae Seung-Hun Lee

Personal detail

Name in Full: Seung-Hun Lee Date of Birth: 16 December 1982 Gender: Male Nationality: Republic of Korea Address: Metroheights 101 Dong 602 Ho, Jeonpo 2-dong, Busanjingu, Busan, Korea Schllighoern 6, Bremen, Germany Telephone Number: +49 176 8464 3086 E-mail: [email protected] , [email protected]

Education Background

Sep. 2012 ~ Ph.D. Analytical Chemistry Advisor: Professor Nikolai Kuhnert Jacobs University Bremen, Germany

Mar. 2005 ~ Feb. 2007 M.Sc. in Organic Chemistry Advisor: Professor Jung-Il Jin Korea University, Korea

Mar. 2001 ~ Feb. 2005 B.Sc. Department of Chemistry Korea University, Korea

Employment

Feb. 2007 ~ Apr. 2010 Researcher Information & Electronic Materials Research Advanced Polymeric Materials R&D Center Samyang Corporation

Certification

2008 Balloon Artist , Party Planner Education Association 2009 Barista , Korea Coffee Education Society 2010 Chinese Character , Korea Test Association 2011 Magic Instructor , The Korean Magic Society 2012 Korean History Proficiency , National Institute of Korean History

Scholarship

2005 1st Semester, BK Scholarship 2005 2nd Semester, General Scholarship 2006 1st Semester, Research Assistant Scholarship 2006 2nd Semester, Administration Assistant Scholarship & Samyang Education Sponsorship 2017 DAAD Merit-based Scholarship (by Wolfgang-Ritter Stiftung)

Activity

2003 Student Leader, Department of chemistry, Korea University 2008 Band Leader, R-band, Samyang Research Center 2009 Mensa Korea Life Member (IQ: over 156 [SD=24]) 2011 Master, Coffle SIG, Mensa Korea

Award

2000 Prize for Encouragement , Math Olympiad, Daejeon, Korea Hannam University & The Korean Mathematical Society 2014 1st Prize , Home Brewers Cup Bremen, Bremen, Germany 2016 Excellent Poster Prize , The Korean Society of Food and Nutrition

Achievement Poster 1. Photopatternable Electroluminescent Polymers: MEH-PPV and PVK 98 th General Meeting of Korean Chemical Society, 2006, Gwangju, Korea 2. Characterization of Three Different Grades of Jamaica Blue Mountain Coffee by LC-MS n Second International Congress on Cocoa Coffee and Tea, 2013, Naples, Italy 3. Analysis of Fermented Coffee Beans as a Roasted Coffee Substitute Bremen Life Sciences Meeting 2015, 2015, Bremen, Germany 4. Identification of Composition Kinetics from Cold-brewed Coffee by LC-MS n Third International Congress on Cocoa Coffee and Tea, 2015, Aveiro, Portugal 5. Identification of Composition Kinetics of Traditional Korean Green Tea Producing Method by LC-MS The Korean Society of Food and Nutrition Summer Conference, 2016, Seoul, Korea 6. Analysis of Less- and Enhanced- Caffeinated Coffee: LeeN-Caffeination Korea Society of Food and Cookery Science, 2017, Seoul, Korea 7. Identification of composition kinetics on applications of the LeeN caffeination coffee Fourth International Congress on Cocoa Coffee and Tea, 2017, Turin, Italy

Publication 1. “Poly(fluorenevinylene) Derivatives by Heck Coupling: Synthesis, Photophysics, and Electroluminescence” J. Polym. Sci. Part A: Polym. Chem. 44 , 4494-4507 (2006). 2. “Photopatternability of poly(vinylcarbazole) bearing cinnamate pendants and its blends with a soluble poly(p-phenylene vinylene) derivative” MACROMOLECULAR RESERCH, 15 , 142 (2007). 3. “Enhanced Lifetime of Organic Light-Emitting Diodes Using an Anthracene Derivative with High Glass Transition Temperature” Journal of Nanoscience and Nanotechnology 13 , 4216-4222 (2013).

4. “Analysis on different grades of highly-rated Jamaica Blue Mountain coffees compared to easily available originated coffee beans” SCIREA Journal of Food. 1(2) , 15-27 (2016) .

Patent 1. Polysilsesquioxane-based organic-inorganic hybrid graft copolymer, organo-silane comprising porogen used for preparation of the same and Method for preparation of insulating film comprising the same, AN=[10-2007 -0140502] 2. Method of preparing insulating film with using norbornene-based silane derivative and Insulating film prepared by the same, AN=[10-2007 -0140517] 3. Norbornene-based silsesquioxane copolymers, Norbornene-based silane derivative used for preparation of the same and Method of preparing low dielectric insulating film comprising the same, AN=[10-2007 -0134577]

4. Photoactive resin composition for insulation layer, AN=[10-2009 -0131139] 5. Positive Photoresist Composition for insulation layer, AN=[10-2009 -0131033] 6. A photosensitive composition for insulation layer, AN=[10-2010 -0130912]

Article in Popular Press 1. Compounds to diabetes in coffee and tea, Coffee & Tea , 2017. 02. 2. Polyphenols and antioxidant effect, Coffee & Tea , 2017. 04. 3. Flavor of coffee? Minimization of Furans?, Coffee & Tea , 2017. 06. 4. Is really Dutch coffee decaffeinated? Are more than 10 % coffee companies unscrupulous?, Coffee & Tea , 2017. 08. 5. Is coffee good for health or not? Why does the debate still go on?, Coffee & Tea , 2017. 10. 6. The leading scientists; who study coffee, Coffee & Tea , 2017. 12. 7. Coffee addiction? Caffeine addiction? Addiction and Dopamine, Coffee & Tea , 2018. 02.

8. Trigonelline and Vitamin B 3 in Coffee, Coffee & Tea , 2018. 04., submitted .

Books 1. Chlorogenic acids in Coffee, Book chapter invited, The Royal Society of Chemistry (RSC)

Research Skills

1. Structure Design, Synthesizing, and Characterization of new materials 2. Analysis on unknown samples (Tools available: NMR, IR, UV, PL, TGA, DSC, GPC, EA, GC-MS, ICP, etc) 3. Interpretation of polyphenols and caffeine with HPLC-TOFMS and HPLC- tandem MS

Interests 1. Food chemistry, analysis of coffee and tea 2. Mass Spectrometry 3. Optoelectronic devices based on organic materials (OLED, OPV, OTFT) 4. Chemical Education

References

Professor Jung-Il Jin Past President, IUPAC Department of Chemistry, Korea University, Korea [email protected]

Professor Nikolai Kuhnert School of Engineering and Science, Jacobs University Bremen, Germany [email protected]

Professor Jeonghun Kwak School of Electrical and Computer Engineering, University of Seoul, Korea [email protected]

Preface

Following water and black tea, coffee is the third most consumed beverage globally with an average daily consumption of 400 mL per capita. Not only in the beverage category, coffee is the third most traded commodity in the category of food and 15 th the entirely of the most valuable commodity in the world.

The world production of green coffee was 3,522,508 tons in 1976 and became 8,790,005 tons in 2014. Especially in China, people started to produce green coffee around 2000, joined the group of coffee producing top 20 countries in the world in 2009, became 12 th coffee producing country in 2013 and still highly increasing their coffee production. Annual coffee trade is worth an annual 120 billion US$ in raw material and coffee provide a livelihood for an estimated 25 million farmers.

Also in the research field, coffee is a very popular and still growing topic. In November, 2012, there were already 38,589 results by SciFinder with the topic of coffee. In August, 2017, the result increased to 52,275, which means an average 3000 papers were published annually related to coffee in the recent five years.

In this respect, to know the constituent of coffee, the processes from farm to table of coffee and coffee analysis methods are needed to be considered together. This project reviews the composition of coffee, processes and analysis methods and shows results of analysis of compositional changes for novel methods of coffee processing through the whole coffee processes.

1. Introduction

1.1. Coffee Chemistry A substantial body of research has supported the notion that drinking coffee is beneficial for human health especially affecting the reduction of cancer, type II diabetes 1, Parkinson’s Disease 2, liver disease 3 and improving athletic performance 4. Hence coffee research has become more and more important. With this flow, the most emphasized compounds in coffee are chlorogenic acids, especially since Clifford et al. developed a new easy and fast analytical method for profiling different regioisomers of chlorogenic acids using HPLC-tandem MS in 2003. 5 After this research, coffee polyphenols, which is represented by chlorogenic acids, entered the mainstream of food research using the innovative method of characterization.

In addition, although there are still controversial arguments addressing its health influence, the compound caffeine is an attractive compound for analysis. 2,4,6,7 Particularly many energy- drinks started to be produced in 21 th century, caffeine became a special issue with respect to daily intake recommendations. Coffee is the main source of caffeine in normal human diet in many countries, so the caffeine research in coffee is one of the two major research fields, together with chlorogenic acids research in coffee. Current figures recommend a daily intake of 400 mg, with consumption of in excess of 1 g / day leading to caffeinism and an LD 50 dose of 150 mg / kg.

In this thesis, the composition of coffee of varying types, chemical compositions and properties are investigated and discussed with a focus on new applications. LC-MS methods are predominantly used for coffee composition analysis.

1.1.1. Coffee Compounds Coffee is mainly composed of the three energy providing substances; carbohydrate, protein and lipids.8 The three substances account for about 80 % of the total dry green coffee beans. Table 1 shows component of coffee in two representative species, Coffea arabica (arabica coffee) and Coffea canephora (robusta coffee) which are the most commonly used coffee products. However, those principal ingredients do not attract people’s interest because we only consume two cups of coffee per day in Germany, which is one of the most coffee consuming countries in the world, and two cups of coffee has only 5 g of solid coffee content.

Table 1. Component of coffee in Arabica and Robusta coffee beans. Components % on dry basis Arabica Robusta Caffeine 1.2 2.2 Trigonelline 1.0 0.7 Minerals (41% = K) 4.2 4.4 Acids Chlorogenic acids – total 6.5 10.0 Aliphatic acids 1.0 1.0 Quinic acid 0.4 0.4 Sugars Sucrose 8.0 4.0 Reducing sugars 0.1 0.4 Polysaccharides 44.0 48.0 Lignin 3.0 3.0 Pectin 2.0 2.0 Proteins 11.0 11.0 Free Amino acids 0.5 0.8 Lipids 16.0 10.0

Although energy providing substances are the main ingredients, the total amount of intake is even less than 1 % of what people usually take from other food. On the other hand, chlorogenic acids and caffeine take up 5 ~ 10 % and 1 ~ 2 %, respectively in the coffee component, but their main source of supply to human is coffee. Therefore, chlorogenic acids and caffeine analysis becomes the main interest on coffee research.

1.1.2. Chlorogenic acids 1.1.2.1. Etymology and classification Chlorogenic acids are natural chemical compounds that are in general terms defined as hydroxycinnamoyl esters of quinic acids.9-11 Typical substituents are caffeic acid, ferulic acid, p-coumaric acid and others. The most common chlorogenic acids in coffee is 5-caffeoylquinic acid of which chemical structure is shown in Figure 1 and it is commercially available in the market with the trivial name of chlorogenic acid. 12

O HO O OH O

HO OH HO OH Figure 1. Chemical structure of 5-caffeoylquinic acid

The name chlorogenic acid comes from the Greek χλωρός (light green) and -γένος (a suffix meaning ‘give rise to’), because of the green color produced when chlorogenic acids are oxidized. 13 The green color obtaining reaction has been found by a group of Japanese scientists, which is by chlorogenic acid oxidizing with protein of sunflower seeds to make green pigments. 14

Chlorogenic acids are classified as one of the polyphenols from the fruit, and these polyphenols are regarded as antioxidant compounds. 15 After the name polyphenol started to be used, research on natural polyphenols became one of the mainstreams in order to explain the benefits of fruits for our health. The ‘French Paradox’ 16 is one of the most famous examples illustrating the potential health benefits of phenolics in cardiovascular diseases. 17

1.1.2.2. Polyphenols Etymology of polyphenol First of all, it should be highlighted that polyphenols are not polymers. Most of the polyphenols do not have repeating units in their structures, and their molecular weights are below 5000 Da. The word polyphenols has coexisted with the terms plant phenolics and tannins until the term polyphenols dominate the literature around 2000.

The most employed definition of polyphenols is WBSSH definition.18 The White-Bate-Smith- Swain-Haslam (WBSSH) describes the polyphenol class as water soluble, 500 – 3000 ~ 4000 Da, possessing 12 to 16 phenolic hydroxyl groups on five to seven aromatic rings per 1000 Da. However, according to this definition, only a few chlorogenic acids might be included in the polyphenol group. 5-Caffeoylquinic acid, which is the most abundant chlorogenic acid in coffee, has only 354.31 g/mol, so that it cannot reach the low limit of 500 Da with the WBSSH definition. Hence CGAs fall under the definition of plant phenolics. It is different from general cognition that coffee has a lot of polyphenolic compounds including caffeoylquinic acids.

Stéphane Quideau suggested that the definition need to be changed in 2011. 15 He proposed a new definition of polyphenol to include the small compounds which could be easily identified after the analytical development of Clifford et al.5 The new definition is shown below.

The term “polyphenol” should be used to define plant secondary metabolites derived exclusively from the shikimate-derived phenylpropanoid and/or the polyketide pathway(s), featuring more than one phenolic ring and being devoid of any nitrogen-based functional group in their most basic structural expression.

In summary, the term polyphenol is now widely used as a common designation of compounds, which contains multiples of phenolic rings.

The role of polyphenols Polyphenol as UV-absorbers It is well-known that the UV-B light cause direct DNA-damage to produce thymine-thymine cyclobutane dimers.19 Since the wavelength of UV-B light is between 290 nm and 320 nm, polyphenols could easily absorb the light. From benzene to phenol, the absorption goes through red-shift from 254 nm to 270 nm. Additional hydroxyl groups produce additional red shift of absorption to the UV-B range. The protection against DNA damaging solar radiation is one of the reasons of mounting research on polyphenols. It is believed that plants when leaving the aqueous habitat, moving on land, have evolved polyphenols as secondary metabolics to serve as UV protecting agents.

Polyphenol as antioxidants Antioxidant property plays a more important role in recent research on polyphenols than DNA protection. In view of several neurological researches such as Alzheimer’s disease or Parkinson’s disease and diabetes, antioxidant study provides important scientific concept in development of drugs for managing the disorders. Polyphenols discovery as antioxidant compounds inspired increment of antioxidant investigation, 20 and seemed to explain well with the health beneficial effect by antioxidant theory. However, the low absorption of the polyphenols into our body weakened the adaption of polyphenol as health beneficial antioxidants. 21 The biological activity of chlorogenic acids would follow to go into details.

1.1.2.3. Chlorogenic acids in coffee Chorogenic acids content in green coffee bean Among the polyphenols, chlorogenic acids have over the last two decades attracted scientific interest. Chlorogenic acids contribute largely to daily dietary burden of polyphenols with an

estimated daily intake of at least 1 g. Coffee is the major CGA source with 280 mg / cup of coffee. A cup is in coffee research defined as a 200 mL volume.

Occurrence Coffee is known as one of the most abundant natural plant sources of chlorogenic acids with bamboo and maté. Although chlorogenic acids are also found in most vegetables and fruits such as peach, prunes and berries 22 , chlorogenic acids research focuses on coffee because coffee is most abundant source of obtaining chlorogenic acids in contemporary human diet. The main compound of chlorogenic acids in coffee is 5-caffeoyl quinic acid. As shown in Figure 2. a) and b) , the three different isomers of caffeoylquinic acids (3-CQA, 5-CQA and 4-CQA) are the main compounds among the coffee composition. Figure 2 shows a typical chromatogram of a coffee infusion. It can be seen that 5-CQA is generally used as the representative standard polyphenolic compound in coffee with the highest abundance in the chromatogram.

Figure 2. LC-MS chromatogram in negative ion mode of a Brazil green coffee extract; a) total ion chromatogram, b) extracted ion chromatogram of m/z 353 ± 0.5

Structures As already mentioned the term chlorogenic acid in chapter 1-1-2. Chlorogenic acids, chlorogenic acids are composed hydroxycinnamic acids esterified with quinic acid. The chemical structure of chlorogenic acids and their nomenclature are shown in Figure 3 and Table 2 below. The priority of carbons in quinic acid is followed by IUPAC nomenclature; Ra, Rc, Rd and Re are the position of 1, 3, 4 and 5, respectively.

O ORe ORd R'O a) ORa ORc

O O O OH O

R’ = b1) OH b2) OH b3) OH Figure 3. Chemical structure of quinic acid (a) ) and typical hydroxycinnamoyl substituents; b1) Caffeoyl-, b2) Feruloyl-, b3) Hydroxycinnamoyl-

Table 2. List of chlorogenic acid derivatives No Name Abb Q Rx R’ 1 Caffeoylshikimic acid CSA 1db2 C H 2 Caffeoyl-1,5-quinide CQL 1la5 C la 3 p-Coumaroylquinic acid CoQA - Co H 4 Caffeoylquinic acid CQA - C H 5 Feruloylquinic acid FQA - F H 6 Caffeoylquinic acid methyl ester M-CQ - C CH 3 7 diCaffeoylquinic acid diCQA - diC H 8 Caffeoylferuloylquinic acid CFQA - C, F H 9 Feruloyltrimethoxycinnamoylquinic acid FTMQA - TM, F H 10 Feruloylsinapoylquinic acid FSQA - S, F H

In Table 3, there are 69 chlorogenic acid compounds structures published by Jaiswal et al pursuant to attached groups in Figure 4.23 These chlorogenic acids in table 3 are all found in natural green coffee beans, and there are also more chlorogenic acids found in roasted coffee formed by a set of well investigated reactions during roasting such as acyl migration, lactonisation, epimerization or dehydration.24,25

Table 3. Reported chlorogenic acids in coffee bean

No. Name Abbreviation Rc Rd Re 1 3-O-caffeoylquinic acid 3-CQA C H H 2 4-O-caffeoylquinic acid 4-CQA H C H 3 5-O-caffeoylquinic acid 5-CQA H H C 4 3-O-feruloylquinic acid 3-FQA F H H 5 4-O-feruloylquinic acid 4-FQA H F H 6 5-O-feruloylquinic acid 5-FQA H H F 7 3-O-p-cormaroylquinic acid 3-pCoQA pCo H H 8 4-O-p-cormaroylquinic acid 4-pCoQA H pCo H 9 5-O-p-cormaroylquinic acid 5-pCoQA H H pCo 10 3-O-dimethoylcinnamoylquinic acid 3-DQA D H H 11 4-O-dimethoylcinnamoylquinic acid 4-DQA H D H 12 5-O-dimethoylcinnamoylquinic acid 5-DQA H H D 13 3-O-sinapoylquinic acid 3-SiQA Si H H 14 4-O-sinapoylquinic acid 4-SiQA H Si H 15 5-O-sinapoylquinic acid 5-SiQA H H Si 16 3,4-di-O-caffeoylquinic acid 3,4-diCQA C C H 17 3,5-di-O-caffeoylquinic acid 3,5-diCQA C H C 18 4,5-di-O-caffeoylquinic acid 4,5-diCQA H C C 19 3,4-di-O-feruloylquinic acid 3,4-diFQA F F H 20 3,5-di-O-feruloylquinic acid 3,5-diFQA F H F 21 4,5-di-O-feruloylquinic acid 4,5-diFQA H F F 22 3,4-di-O-p-cormaroylquinic acid 3,4-di pCoQA pCo pCo H 23 3,5-di-O-p-cormaroylquinic acid 3,5-di pCoQA pCo H pCo 24 4,5-di-O-p-cormaroylquinic acid 4,5-di pCoQA H pCo pCo 25 3-O-feruloyl-4-o-caffeoylquinic acid 3F-4CQA F C H 26 3-O-caffeoyl-4-o-feruloylquinic acid 3C-4FQA C F H 27 3-O-feruloyl-5-o-caffeoylquinic acid 3F-5CQA F H C 28 3-O-caffeoyl-5-o-feruloylquinic acid 3C-5FQA C H F 29 4-O-feruloyl-5-o-caffeoylquinic acid 4F-5CQA H F C 30 4-O-caffeoyl-5-o-feruloylquinic acid 4C-5FQA H C F 31 3-O-dimethoxycinnamoyl-4-o-caffeoylquinic acid 3D-4CQA D C H 32 3-O-dimethoxycinnamoyl-5-o-caffeoylquinic acid 3D-5CQA D H C 33 4-O-dimethoxycinnamoyl-5-o-caffeoylquinic acid 4D-5CQA H D C 34 3-O-dimethoxycinnamoyl-4-o-feruloylquinic acid 3D-4FQA D F H 35 3-O-dimethoxycinnamoyl-5-o-feruloylquinic acid 3D-5FQA D H F 36 4-O-dimethoxycinnamoyl-5-o-feruloylquinic acid 4D-5FQA H D F 37 3-O-p-coumaroyl-4-o-caffeoylquinic acid 3pCo-4CQA pCo C H 38 3-O-caffeoyl-4-o-p-coumaroylquinic acid 3C-4pCoQA C pCo H 39 3-O-p-coumaroyl-5-o-caffeoylquinic acid 3pCo-5CQA pCo H C 40 3-O-caffeoyl-5-o-p-coumaroylquinic acid 3C-5pCoQA C H pCo 41 4-O-caffeoyl-5-o-p-coumaroylquinic acid 4C-5pCoQA H C pCo 42 4-O-p-coumaroyl-5-o-caffeoylquinic acid 4pCo-5CQA H pCo C 43 3-O-p-coumaroyl-4-o-feruloylquinic acid 3pCo-4FQA pCo F H 44 3-O-p-coumaroyl-5-o-feruloylquinic acid 3pCo-5FQA pCo H F 45 4-O-p-coumaroyl-5-o-feruloylquinic acid 4pCo-5FQA H pCo F

46 4-O-dimethoxycinnamoyl-5-o-p-coumaroylquinic acid 4D-5pCoQA H D pCo 47 3-O-p-coumaroyl-4-o-dimethoxycinnamoylquinic acid 3pCo-4DQA pCo D H 48 3-O-p-coumaroyl-5-o-dimethoxycinnamoylquinic acid 3pCo-5DQA pCo H D 49 3-O-sinapoyl-5-o-caffeoylquinic acid 3Si-5CQA Si H C 50 3-O-sinapoyl-4-o-caffeoylquinic acid 3Si-4CQA Si C H 51 3-O-(3,5-dihydroxy-4-methoxy)cinnamoyl-4-o-feruloylquinic acid 3DM-4FQA DM F H 52 4-O-sinapoyl-3-o-caffeoylquinic acid 4Si-3CQA C Si H 53 3-O-sinapoyl-5-o-feruloylquinic acid 3Si-5FQA Si H F 54 4-O-sinapoyl-5-o-feruloylquinic acid 4Si-5FQA H Si F 55 4-O-sinapoyl-3-o-feruloylquinic acid 4Si-3FQA F Si H 56 4-O-trimethoxycinnamoyl-5-o-caffeoylquinic acid 4T-5CQA H T C 57 3-O-trimethoxycinnamoyl-5-o-caffeoylquinic acid 3T-5CQA T H C 58 3-O-trimethoxycinnamoyl-5-o-feruloylquinic acid 3T-5FQA T H F 59 3-O-trimethoxycinnamoyl-4-o-feruloylquinic acid 3T-4FQA T F H 60 4-O-trimethoxycinnamoyl-5-o-feruloylquinic acid 4T-5FQA H T F 61 3-O-dimethoxycinnamoyl-4-o-feruloyl-5-o-caffeoylquinic acid 3D-4F-5CQA D F C 62 3,4,5-tri-O-caffeoylquinic acid 3,4,5-triCQA C C C 63 3,5-di-O-caffeoyl-4-o-feruloylquinic acid 3,5-diC-4FQA C F C 64 3-O-feruloyl-4,5-di-o-caffeoylquinic acid 3F-4,5-diCQA F C C 65 3,4-di-O-caffeoyl-5-o-feruloylquinic acid 3,4-diC-5FQA C C F 66 3-O-caffeoyl-4,5-di-o-feruloylquinic acid 3C-4,5-diFQA C F F 67 3,4-di-O-feruloyl-5-o-caffeoylquinic acid 3,4-diF-5CQA F F C 68 3,4-di-O-caffeoyl-5-o-sinapoylquinic acid 3,4-diC-5SiQA C C Si 69 3-O-sinapoyl-4,5-di-o-caffeoylquinic acid 3Si-4,5-diCQA Si C C

O O O HO OH HO OH HO OH O O HO OH O OH HO OH HO O OH O O OH HO OH

OH OH 1 2 3

O O O HO OH HO OH HO OH O O O O O OH HO OH HO O OH O O OH HO OH

O OH 4 5 6

O O O HO OH HO OH HO OH O O

O OH HO OH HO O OH O O OH HO OH

OH 7 8 9

Figure 4. Chemical structures of chlorogenic acids in Table 3 . (Continues 1/8)

O O O HO OH HO OH HO OH O O O O O OH HO OH HO O OH O O OH O O

O O 10 11 12

O O O HO OH HO OH HO OH O O O O O OH HO OH HO O OH O O OH HO OH O O

O O OH 13 14 15 O O HO OH HO OH O O O HO O OH HO O O O O O H HO OH HO O H

OH OH 16 17 O O HO OH HO OH O O O OH O OH HO O O O O O HO OH

O OH OH OH 18 19 Figure 4. Chemical structures of chlorogenic acids in Table 3 . (Continues 2/8) 30

O O HO OH HO OH O O O O O O HO O O O O O OH OH HO OH

O OH 20 21 O O HO OH HO OH O O O O OH O O O O HO OH HO O H

OH 22 23 O O O HO OH HO OH HO OH O O O O HO O OH O OH HO O O O O O O O HO HO OH

OH O OH OH OH 24 25 26 O O HO OH HO OH O O O O O OH HO O O O O O OH OH HO OH HO OH

27 28 Figure 4. Chemical structures of chlorogenic acids in Table 3 . (Continues 3/8)

O O O HO OH HO OH HO OH O O O O OH O O OH HO O HO O O O O O O O O OH OH

OH O OH OH OH OH 29 30 31 O HO OH O O OH HO OH HO O O O O O OH O OH O O OH O OH

O O 32 33 O O HO OH HO OH O O O O O OH O O O O O O O OH O OH

O OH 34 35 O O O HO OH HO OH HO OH O O O O HO HO O O OH O OH O O O O O O OH HO HO

O OH O OH OH 36 37 38 Figure 4. Chemical structures of chlorogenic acids in Table 3 . (Continues 4/8)

O O HO OH HO OH O O O O OH HO O O O O OH OH HO OH HO OH 39 40 O O O HO OH HO OH HO OH O O O OH HO O HO O O OH O O O O O O OH OH HO

OH O OH OH OH 41 42 43 O HO OH O O O HO OH HO O O O O O OH O O O OH HO OH

44 45 O O HO OH HO OH O O

HO O O OH O O O O OH HO

O O O O 46 47 Figure 4. Chemical structures of chlorogenic acids in Table 3 . (Continues 5/8)

O O HO OH HO OH O O O O O O OH O O O O OH OH HO O HO OH O 48 49

O O O HO OH HO OH HO OH O O O O HO O O OH O OH O OH O O O O O O HO O HO O OH O

OH O OH OH OH OH 50 51 52 O HO OH O O O HO OH HO O O O O O OH O O O O OH HO OH O O O OH

53 54 O O HO OH HO OH O O O OH O OH HO O O O O O HO OH

O O O O OH O 55 56 Figure 4. Chemical structures of chlorogenic acids in Table 3 . (Continues 6/8)

O O HO OH HO OH O O O O O OH O O O O O O OH OH O OH O OH O O 57 58 O O HO OH HO OH O O O O O OH HO O O O O O O OH O

O O O OH O 59 60 O O HO OH HO OH O O O O O OH HO O O O O O H O O O O O OH HO O H

O OH OH OH 61 62 O O HO OH HO OH O O O O HO OH O O O O O O H O O O O HO OH HO O H

O OH OH OH 63 64 Figure 4. Chemical structures of chlorogenic acids in Table 3 . (Continues 7/8)

O O HO OH HO OH O O O O HO O HO O O O O O O O O O HO OH HO OH

OH O OH OH 65 66 O O HO OH HO OH O O O O O OH HO O O O O O O O O O HO OH HO OH O

O OH OH OH 67 68 O HO OH O O O OH O O O O HO OH O

OH OH 69 Figure 4. Chemical structures of chlorogenic acids in Table 3 . (Continued 8/8)

1.1.2.4. Analysis of chlorogenic acid Analytical methods of chlorogenic acids Although it is possible to analyze chlorogenic acids with a variety of different analytic tools such as 13 C-NMR, 1D 26 and 2D-1H-NMR 27 , FT-IR 28 , FT-Raman 29 and so on, the major method to analyze chlorogenic acids is HPLC-MS. In recent studies, the comparison of those spectroscopic techniques for analyzing chlorogenic acids in coffee 30 is described.

The most common method for analyzing coffee chlorogenic acids was to compare synthesized standard compounds using HPLC-UV/Vis spectra. Concentrations of the CQA isomers and caffeine in weight percent from different originated coffee beans were reported by Michael P. Purdon and David A. McCamey in 1987 as shown in Table 4.a) Concentrations and 4.b) Information of the coffee beans .

Table 4.a) Concentrations of the CQA isomers and caffeine in weight percent

Table 4.b) Information of the coffee beans used in Table 4(a)

In 2003, different fragmentation patterns of three different caffeoyl quinic acids were reported. 5 CGA of regioisomeric MS fragmentations did not attract scientists’ interest because regioisomers usually give the same fragments. However, this report showed clearly different MS 2 spectra, allowing unambiguous assignment of regiochemistry, so that HPLC-tandem MS became the most important and popular tool for analyzing chlorogenic acids in coffee extracts. 37

Underestimated solubility differences for the real coffee consumption Meanwhile, analyzing coffee started from the view of analyzing food, coffee beans were extracted with 70 % methanol solutions using common established methods for the extractions of food, vegetables and fruits. Although we consume all the coffee beans in some brewing methods like Turkish coffee, we generally extract only 10 ~ 15 %, 15 ~ 25 % and 45 ~ 60% of water soluble part of coffee for drinking with different methods of hand-brewing, espresso and instant coffee, respectively. Therefore, there were always some differences between research results and real consuming compounds. The differences were studied, 5- CQA equivalent peak areas were measured and would be discussed in this thesis.

1.1.2.5. Properties of chlorogenic acid Organoleptic properties Sweet, acidic, bitter, body, aroma, after taste and green flavor properties are usually discussed with regards to coffee taste. The chlorogenic acids compounds in Table 3 provide complex taste perceived in coffee, together with sweetness in sugars, bitterness in caffeine, body in lipid and the aroma providing volatile compounds that could be identified by GC-MS.31 Generally, the well-known compound suspected to be responsible for the bitterness is caffeine and diketopiperazines (DKPs).32,33 Caffeine and DKPs have been studied for taste research as a representative of bitter compounds in coffee, but a variety of chlorogenic acids plays key role in bitter taste of coffee as well.

Including caffeine and DKPs, a variety of bitter compounds in coffee were identified.34 Not only different regioisomers of caffeoylquinic acids, but also caffeic acid, di-caffeoylquinic acids and caffeoylquinic acid lactones 27,35 play substantial roles in the bitterness of coffee. Although there are a lot of bitter-tasting molecules in coffee, those 10 chlorogenic acids; 3-O- caffeoyl-γ-quinide (1), 4-O-caffeoyl-γ-quinide (2), 4-O-caffeoyl-muco-γ-quinide (3), 5-O- caffeoyl-muco-γ-quinide (4), 5-O-caffeoyl-epi-δ-quinide (5), 3-O-feruloyl-γ-quinide (6), 4-O- feruloyl-γ-quinide (7) 3,4-O-dicaffeoyl-γ-quinide (8), 4,5-O-dicaffeoyl-muco-γ-quinide (9) and 3,5-O-dicaffeoyl-epi-δ-quinide (10); take important parts in including bitterness perception as reported by the Hofmann group.36

As already mentioned in chapter 1-1-1. Coffee compounds, Robusta coffee has a higher caffeine and chlorogenic acid content if compared to Arabica coffee. This is the reason why

sensory panels assign a higher degree of bitterness to Robusta beans. Since bitterness is a less desired attribute in coffee, Arabica beans achieve higher prices in the market.

Biological activity As discussed in chapter previously, chlorogenic acids may play an important role in biological activity as an antioxidant. For a chemist, an antioxidant is a reducing agent. Since reactive oxygen species concept based antioxidant theory was simple and well-explainable both chemically and practically 37,38 , it has become so successful to be accepted by the public as a general truth that antioxidant containing food are beneficial to health.

From the starting of coffee research, the biological activity research of chlorogenic acids only focused on their antioxidantial properties. This flow was due to the view of chlorogenic acids as an extension of polyphenols. Most polyphenol researches were focused on the usage of polyphenols as an antioxidant, and chlorogenic acids share the common antioxidant reactivity of polyphenols.

However, over the last decade, doubts of the antioxidant concept have emerged from various concerns.21 Most dietary antioxidants show an even lower plasma concentration than ascorbic acid or tocopherols and most polyphenols undergo extensive, mostly reductive gut floral metabolism so that produced less efficient antioxidants.39 Although these evidences might weaken the strong position of coffee as a healthy food, these also give the driving force of new research not relying on only the simple explanation of antioxidant as a panacea.

As coffee market grows, several works on the bio-activity of coffee have been documented. The coffee research which includes not only antioxidant properties, but also the antibacterial, antimutagenic and antiviral activities could be correlated with chlorogenic acids.40,41 These studies on how coffee components alleviate the effects of diseases are related to caffeine 42 and chlorogenic acids studies 43 .

1.1.3. Caffeine 1.1.3.1. Etymology and classification Caffeine was discovered by a German chemist, Friedlieb Ferdinand Runge in 1819 at the request of Johann Wolfgang von Goethe. 44 At this time, Runge suggested the name ‘Kaffebase’ that means a base that exist in coffee. However, a French chemist Pierre Jean

Robiquet made the first oral announcement about it at the meeting of the Pharmacy Society in Paris with the name of ‘Caféine ’, meaning amine (-ine) in coffee in French ‘Café ’ and it became generally known as caffeine.

Caffeine is a natural methylxanthine alkaloid, chemically related to adenine and guanine. The chemical structure of caffeine is shown in Figure 5. And for the toxicity, it is classified as ‘generally recognized as safe’ by the US Food and Drug Administration.

N N

O N N

Figure 5. Chemical structure of caffeine

It is known that the individual sensitivity is genetically determined and leads to varying 45 individual caffeine sensitivity. The LD 50 value of caffeine in human is an estimated value of 150 ~ 200 mg per kg, and it means almost two cups of coffee per kg. With continuous overdose, more than 1 g per day per person can cause a physiological effect, so called ‘caffeinism’, including nervousness, irritability, restlessness, insomnia, headaches, palpitations. 46 Therefore, the daily recommended amount of caffeine in most countries is set 400 mg per day for normal adults and 300 mg per day for pregnant woman. An average cup of coffee contains around 100 mg caffeine. Overdosage of coffee powder can, however, lead to commercial cups as high as 323 mg. 47

Coffee is the main source for caffeine in 21 st century’s European and some Asian countries’ diet. It has both positive and negative health effects. By this reason, caffeine becomes the most attractive compound in coffee and scientists produces a lot of results related to caffeine in coffee.

1.1.3.2. The role of caffeine It is speculated that plants are making caffeine to protect themselves against predator insects and to prevent germination of nearby seeds. To humans, it is the most remarkable compound in coffee that causes both positive and negative health effects. People started to use coffee as stimulant drug and this stimulant effect is originated from the caffeine. From twentieth 40

century, people started to make energy drinks containing doses of caffeine of 6 ~ 242 mg / L to use this stimulant effect of caffeine. 48 However, by increasing coffee, chocolate and caffeine-containing energy drink consumption, negative health effect of the caffeine came to the fore.

Since caffeine is chemically one of the methylxanthine families, it interferes adenosine receptors in our brain. By this effect, caffeine intake plays a stimulant role in our body, together with inducing insomnia. It is reported that the caffeine intake reduces arteriosclerosis, hepatic cirrhosis, depression, increasing concentration and agility, diuresis and preventing cancer and obesity. 49 On the other hand, it is also reported that the caffeine intake prevents absorption of iron and calcium, triggers anxiety and hyperacidity, relaxes bronchial muscle and makes addiction. 50 These physiological effect and outburst of intake on caffeine- containing food made caffeine as an attractive target compound for research.

The European Food Safety Authority reported ‘Scientific Opinion on the safety of caffeine’ in 2015. 51 In this report, for adults in the general population, caffeine intakes up to 400 mg per day do not raise safety concerns. For pregnant women, 200 mg per day do not raise safety concern for the fetus. Before this report, most countries recommended 400 mg per day for adults and 300 mg per day for pregnant or breast feeding women followed by the report from Health Canada in 2012. The recommended 400 mg of caffeine is about 5 cups of coffee. The

LD 50 value of caffeine is between 150 ~ 200 mg / kg, it means almost 100 cups of coffee for an adult.

1.1.3.3. Coffee decaffeination In 1903, Ludwig Roselius invented a coffee decaffeination method in Bremen, Germany. 52 It was the first chemical application used in coffee processing. Decaffeination was the first and the last caffeine control method, which could reduce caffeine by more than 97%. At the beginning, Roselius used benzene for caffeine extraction as a solvent, which is now banned as carcinogenic chemical compound. Therefore, there have been a lot of different challenges for extracting caffeine from green coffee beans using methylene chloride, chloroform, ethyl acetate, supercritical carbon dioxide and water. These decaffeination methods will be discussed in following chapter coffee technology, green coffee, decaffeination.

1.2. Coffee Technology From farm to table, coffee changes from the coffee cherry to green coffee to roasted coffee to ground coffee to a final cup of coffee. In each step, many different methods are used and all processes influence the coffee flavor to yield different grades of coffee. In this chapter, each step would be described and the property changes by the processes would be discussed to use for upgrading processes in health beneficial composition based control.

1.2.1. Coffee Cherry 1.2.1.1. Classifications Commercial coffee is composed of three species in the genus Coffea . Coffea Arabica (Arabica coffee) accounts for about 60 % of the total production while Coffea canephora (Robusta coffee) accounts for about 40 %. The third species is Coffea Liberica , which is now less than 1 % of production. (ICO Trade statistics tables, 2015) With the demand of low cost from the global coffee companies, robusta coffee production has increased, but recently, with the demand of specialty coffee, arabica coffee production is increasing again. (ICO Trade statics tables, 2015 – 2016)

After the species, origin and grades are the main price-determining factors. Coffee is cultivated only in the tropical region, referred to as beanbelt or coffeebelt, shown in Figure 6. There are three main different criteria for determining coffee grades. Height of the cultivation, size of the coffee bean and the number of defected beans are the three criteria. The application of the standards depends on the country; Tanzania and Kenya by size, Ethiopia and Indonesia by the number of defected beans, Guatemala by height of the cultivation, Hawaii by Size and defected beans and Jamaica by size and height. Together with the main three criteria, there are some special standards such as animal processed (Indonesia Luwak), wind fermented (India Monsoon) or Peaberry (Kenya PB).

Figure 6. Coffee Belt / Bean Belt (Figure from Wikipedia)

1.2.1.2. Processes After harvesting the coffee cherry, the next step includes production of green coffee bean by so called ‘Coffee Process’. The aim of these processes is to remove the flesh of the cherry to yield the seed referred to as coffee bean. In general, there are three different processes in this step; wet, dry and semi-dry processes.

Dry process is a predominant method because it was the oldest method and do not need high cost of equipment. Two third of world coffee production is conducted by the dry process. Furthermore, after 2010, coffee with a variety of fermentation methods become popular. The dry process was preferred and could be easily adapted for fermentation. Almost all the robusta coffee processes and a major portion of arabica coffee in Brazil, Ethiopia, Haiti, Paraguay, India and Ecuador are dry process. In the dry process, coffee cherries after harvested are placed in the sun on tables or thin layers on the ground or patios. It takes some days to max. 4 weeks to dry the cherries and they are hulled to make green coffee beans.

Wet process have used in the countries which are not suitable for dry process because of frequent rainfalls; Vietnam, India and Indonesia. This process needs cost of equipment, but gives a cleaner product. Therefore, washed-bean or full-washed bean had started to become popular especially in the highly-evaluated origin coffee regions. However, with the fame of the fermentation coffee, dry process became again preferred. Although wet process is evolved

into ferment-and wash method, still it is used in limitative origins. In the wet process, the fruit covering the beans is removed before they are dried. The harvested cherries are immersed into water to sort out the unripe fruit by floating. The sunk cherries are flown by the water flow to be hulled with mechanical demucilaging and dried to obtain green coffee beans.

Semi-dry process, also called wet-hulled, semi-washed or pulped natural, is a hybrid process used in several farms of Indonesia and Brazil. This method removes the outer skins by dry- process and removes mucilage by wet-process.

There are research results investigating different methods, but no published article directly compares coffee beans of identical origin with different processes. This is because all the processes are selected by its facilities, but no coffee farm equipped different process facilities together. Therefore, there are still under investigated research topics, which suffer from difficulties in sample acquisition.

1.2.2. Green Coffee 1.2.2.1. Classifications There are some special coffee beans processed prior to yielding the green coffee bean. These special treatments are mostly used for improving medicinal properties or are based on traditional believes, all associated with chemical changes in coffee composition. Decaffeination, fermentation and animal processed coffees are the results of such special treatments.

Normal coffee This is the standard of comparison. Normal coffee is processed by one of the coffee processes mentioned in 1-2-1-2. Processes without further treatment. By using decaffeination or fermentation, the normal coffee can change into special coffee.

Special coffee Special coffee could be obtained by applying a special method both before and after the normal coffee is obtained. By using the process-completed normal green coffee bean, decaffeination or fermentation could be applied to change the chemical compositions for special use. Besides, animal processed coffees could be obtained simply by giving the coffee

cherry to the animals. Presently, the animal processed coffee without going through the normal green coffee process is the most expensive coffee in the world.

Decaffeination The first chemical application and the first caffeine control method in coffee is decaffeination. It had been more than 80 years from the first isolation of caffeine from coffee to the first trial of removing caffeine from coffee. Ludwig Roselius obtained decaffeinated coffee by steaming or using benzene to extract caffeine from coffee in 1903. This method is patented in 1906 and he founded Kaffee HAG in Bremen to produce the decaffeinated coffee marketing it with a slogan of stomach-friendly coffee. However, the method is stopped now because it was proven that the benzene is carcinogen. Therefore, a variety of methods are developed.

Methylene chloride was the first substitute for benzene because of its good extracting character and low boiling point for removing after extracting caffeine without known harmful effect on health. However, methylene chloride is now on the list of International Agency for Research on Cancer (IARC) in Group 2A, which means it is probable carcinogen. For this reason, many coffee companies have banned this method, but there are still some global companies using methylene chloride method. 53

Also, ethyl acetate was used as the substitute for benzene and it is still accepted because this solvent is relatively safe (LD 50 for rat is 5620 mg / kg) and present as a natural constituent in fruits, so this method is called as natural process. 54 However, this method also needs chemical solvent and has no other advantages compared to other methods but using the name ‘natural’. Therefore, the method using ethyl acetate could not occupy the major market of decaffeination methods.

Supercritical fluid extraction of caffeine is one of the most widely used processes, which use 55 supercritical CO 2 as the extraction solvent. This method was invented by Kurt Zosel from the Max Planck Institute for Coal Research in Mülheim an der Ruhr in 1967 and patented three years after the development. The only disadvantage of this method is the high cost of equipment and maintenance along with co-extraction of bioactives in particular CQA-lactones.

The last decaffeination is so called Swiss Water Process. 56 This method uses water as extracting solvent and this is why this method could become popular in relatively short time.

First, coffees extracted with water, filtered with charcoal to remove caffeine, and the green coffee extract without caffeine are used to extract only caffeine from other batch of coffees and recycled the method from charcoal filtering.

Coffee fermentation Coffee fermentation is an old terminology, which is used for the short storage period during the coffee process. No matter which process among the wet, semi-dry or dry process was selected, they all have a small period for the coffee beans to be exposed with the residues of mucilage. During this period coffee beans are exposed to microorganisms. After 2000, specialty coffee becomes a trend, people started to improve the fermentation method for ameliorating its flavor. A variety of controls have been done with temperature, humidity, sun dry or additive controls, not only for the traditional coffee fermentation periods but also after green bean or roasted bean steps. Furthermore, there were natural fermentation processes by using seasonal wind or animal organs. Especially animal process would be dealt with following chapter because it has the character not only with the fermentation, but also animal selection.

Fermentation on coffee process As specialty coffee becomes more common, the coffee industry tends towards developing new methods for making coffee beans to be more special. Fermentation control is one of the efforts conducted from the farm in this regard. Coffee fermentation is already an established step during coffee process, though there are differences in their detailed methods. From wet to dry process, the fermentation period is increased so that the dry process is easily adapted by the fermentation method. Coffee producers started to control temperature, humidity, sunlight exposure and the use of some bacteria on the coffee beans. The controlled fermentation method has been improved and finally in 2015, these fermentation controlled coffee bean became the main source of 13 th Korea Barista Championship champion. These enhanced fermentation coffee is known to improve their fruit flavor in coffee beans.

Fermentation after green coffee bean In 2011, a research on the use of steaming and drying of coffee was published as a patent which indicates that this special treatment of green coffee bean changed green flavor to fruit flavor, as well as the noble processed coffee bean extracted made it aroma richer than the reference-roasted coffee bean extracted by sensory evaluation. 57 Although this method is hard

to include as fermented bean because it did not intended to use microorganisms, it did not prevent from the activities of their inherent microorganisms. However, there is no in-depth investigation for this novel method processed coffee beans because it is a very recent try and succeeded in a limited scale of production.

Fermentation by seasonal wind There is a special yellow green bean in India. The coffee is fermented green coffee beans by the seasonal wind, ‘Monsoon’. During the times of British Raj, Indian coffee beans were transported to Europe by sea for three to four months. The humidity and monsoon wind caused the green coffee to producing a pale yellow color. However, the transportation ways have been improved; it reduced the exposed time for the green coffee beans to the monsoon wind. So the beans do not ripen enough as they were, and people started some artificial fermentation for the coffee beans to replicate the old method, which can give a special soft coffee flavor. However, since it is only produced in the Malabar coast region, it did not receive attention from scientists. Therefore, there is very few published research results about monsooned coffee. 58

Animal processing The most expensive coffee in the world was Kopi Luwak from Indonesia and changed to Black Ivory from Thailand. The two different coffees have one thing in common. They are animal processed coffees. Kopi Luwak is the coffee eaten by civets, and Black Ivory is by elephants. Animals feed on coffee cherries and undigested coffee beans are gathered from the animals’ excrement, washed to become the commercial green coffee bean product.

Kopi Luwak has two advantages over normal coffee with fermentation and animal selection. By passing through the civet’s organs, coffee bean is exposed to protease enzymes of the civet, to make shorter peptides and more free amino acids. In 2013, research on short chain fatty acids compared Kopi Luwak to normal coffee and showed that Kopi Luwak has higher malic acid and citric acid than normal coffee. 59 Also there is so-called animal selection, which means that animals have sixth-sense to figure out which fruit is well-ripened to eat. Therefore, the coffee beans in the excrement of animals are qualified coffee beans by the animal selection.

Black Ivory Coffee claims its advantage over Kopi Luwak that elephants are herbivores in contrast with civets, which are omnivores. They both use their digestive enzymes to break down the coffee’s protein, but herbivores utilize fermentation to help break down cellulose. So the Black Ivory can get smoother taste.

Until now, Black Ivory Coffee was not analyzed and published by scientists because of its high price and low production. However, to compare different animal processed coffees such as civet from Indonesia and Philippines, elephant from Thailand and weasel from Vietnam would be an interesting further research.

1.2.2.2. Processes Excepting only very few special treatment on green coffee bean, the process on green coffee is roasting. Including 2-furfuryl thiol which has high odor activity value in coffee, the compounds, which contribute to ‘coffee flavor’ in technical term, are made by roasting. 60,61 That is to say that roasting is an essential process for coffee flavor. Concentration and extraction yield of major coffee aroma compounds and their chemical structures are shown in Table 5 and Figure 7, respectively. The detailed coffee volatiles with threshold concentration would be discussed later in 1.3. Coffee Sensory Perception.

Table 5. Concentrations of the odorants in coffee powder and coffee brew, and extraction yields of the compounds 60

Concentration Extraction Nr. Compound Powder ( µg/kg) Brew ( µg/l) yield 1 Acetaldehyde 120,000 4,700 73 2 Methylpropanal 24,000 760 59 3 2-Methylbutanal 26,000 870 62 4 3-Methylbutanal 17,000 570 62 5 2,3-Buranedione 49,000 2,100 79 6 2,3-Pentanedione 35,000 1,600 85 7 2-Ethyl-3,5-dimethylpyrazine 400 17 79 8 2-Ethenyl-3,5-dimethylpyrazine 53 1.0 35 9 2,3-Diethyl-5-methylpyrazine 100 3.6 67 10 2-Ethenyl-3-ethyl-5-methylpyrazine 15 0.2 25

11 3-Isobutyl-2-methoxypyrazine 120 1.5 23 12 4-Hydroxy-2,5-dimethyl-3(2 H)-furanone 140,000 7,200 95 13 2-Ethyl-4-hydroxy-5-methyl-3(2 H)-furanone 16,000 800 93 14 3-Hydroxy-4,5-dimethyl-2(5 H)-furanone 1,900 80 78 15 (E)-β-Damascenone 260 1.6 11 16 Guaiacol 2,400 120 65 17 4-Ethylguaiacol 1,800 48 49 18 4-Vinylguaiacol 45,000 740 30 19 Vanillin 4,100 210 95 20 2-Furfurylthiol 1,700 17 19 21 Methional 250 10 74 22 3-Mercapto-3-methylbutyl formate 130 5.7 81 23 2-Methyl-3-furanthiol 60 1.1 34 24 3-Methyl-2-buten-1-thiol 13 0.6 85 25 Methanethiol 4,400 170 72

O O O O O O

1 2 3 4 5 O N N N N O N N N N

6 7 8 9 10

O HO N O HO O O O O N O O OH 11 12 13 14 15

HO HO HO HO

O O HS O O O O

16 17 18 19 20

HS SH SH O S O O SH O

21 22 23 24 25 Figure 7. Chemical structures of major coffee aroma compounds in Table 5.

After reporting the significant decrement of chlorogenic acids content during roasting stage, 62 people started to find a better way to consume more chlorogenic acids in the source of coffee. There was a green coffee extract study in 2011 which showed that regular drinking of green coffee extract can reduce weight and it was suggested that the reason is possibly related to the richness of the chlorogenic acids in green coffee extract. 63 However, participants in the studies were instructed to restrict their diet and increase their exercise in addition to taking the supplement. In 2012, Dr. Oz Show broadcasted this and green coffee extract became famous as a weight loss supplement until Federal Trade Commission fined the sponsor company in 2014. 64

Although the magical weight loss by the green coffee extract was revealed as a happening, the fact that chlorogenic acids undergo decomposition during roasting coffee became a common knowledge despite of its short publication time. And the fact that chlorogenic acids can help losing weight was proven as true through blocking sugar transfer by chlorogenic acids. 65 So the green coffee extract market formed by this happening is still growing.

Not only just removing the roasting step from the coffee process, but also trying to substitute the roasting step was conducted. It is adapting Korean traditional red-ginseng process into green coffee bean to make fermented coffee bean. The process and chemical analysis would be described in this thesis.

1.2.3. Roasted Coffee In roasting process, direct, indirect and semi-direct fire roastings were used. Recently, electricity only, without fire roaster appears with the advantage of reproducibility, represented by Stronghold. Although the type of roasting process plays an important role in coffee flavor, it is hard to examine because the cost of equipment is high for roasters. Likewise, the different classifications of roasting degree and processes after roasting would be discussed in this research.

1.2.3.1. Classifications Excepting some special methods such as steamented coffee and green coffee extract, coffee roasting is an essential process prior to forming basis of coffee brew. Coffee needs different roasting conditions depending on their species, origins and methods of brewing. Recently, the relations of various different roasting conditions and tastes are started to be researched. However, traditional classifications of roasting degree were determined by both roasting time and temperature, changed by 16 level systems by Specialty Coffee Association America (SCAA) standard, and finally expressed by the results of colorimetric measurements.

Traditional classifications The basic categories of traditional roasting degree are green, light, medium and dark roast. Among them, cinnamon, city, full city, French and Italian roast are used, without precise categorization. Generally, the coffee industry used more common names and scientists used more degree names. These traditional classifications are still used both industry and academia as shown in Figure 8, the research of coffee roasting in 2009, 62 but step by step Colorimeter classification substitutes for the traditional classifications. According to the traditional classifications, cinnamon roast is the period for 1st crack and full city roast is the period for 2nd crack. Cinnamon roast to city roast are used for filter coffee brewing and higher than full city roasting are used for espresso brewing.

Figure 8. Total chlorogenic acids found in various green coffee beans and in various roasted coffees treated under different conditions 62

SCAA 16 step classification SCAA tried to change the inconsistent traditional roasting degrees because all roasters have their own roasting process in their separate ways of roasting time and temperatures. SCAA selected not to take account of the process, but to focus only for the resulting color. They divided coffee color into 16 levels from green coffee beans to fully-roasted coffee beans and published as the color-standard for comparing.

Colorimeter classification There is limitation in print-out color grading in human perception. The textile of coffee bean, the type of papers for standard color printout and even the feeling of the roaster can affect for the color perception. To correct this, people started to use colorimetric instruments. The first accepted was Agtron analytic instrument, which has selected by SCAA, for that reason, its analytical value symbolizes the coffee roasting degree. After coffee industry accepts the machine value, HunterLab Coffee Color Index (HCCI) value started to show the degree of roasted coffee. This HCCI value measure the absorbance of 640 nm wavelength, which is orange-red, and calculate the value with the slope of the UV-Visible spectra. Therefore, this HCCI value can present not only the absolute darkness of the coffee bean, but also color changes. However, the relation between the two different machine values and roasting condition was not studied. In contrast most other foods are subjected to colorimetric measurements using the L, a and b scale, coffee colorimetric measurements are using only limited range of wavelengths because the color of coffee beans are limited in yellow green to dark brown during the roasting period. In general, roasted coffee beans have L, a and b values from 30, 14, 23 on light roasting to 12, 4 and 3 on dark roasting, respectively.

1.2.3.2. Processes Storage Coffee is scientifically known as a shelf-stable product after roasting that it does not spoil by enzymatic or microbial processes. However, in sensory tests, a drastic taste change in rancidity is detected, which is triggered from the roasting. These taste changes are originated from either the chemical reactions by the conditions of temperature, humidity and oxygen or the physical diffusion.

The flavor perception is the combination of taste and smell. And in specialty coffee, smell plays more roles. Therefore, though the major compounds of roasted coffee do not change, only small part of the composition changes can make the flavor totally different especially if the small part is volatile. During the roasting process, Strecker degradation makes CO 2, and this gives pressure to break the bean shell for the other volatile compounds to be released from the roasted coffee beans. And lipid oxidation produces odor active aldehydes, accelerated by temperature, humidity and oxygen atmosphere.

For preventing those rancidity sources, coffee roasters use vacuum, nitrogen-charged or one- way valve packages for coffee storage.

Grinding After roasting, grinding should be followed for extraction yield before brewing. It is always recommended that the grinding should be right before coffee brewing because it increases surface area to accelerate staling.

For grinding coffee, two types of grinders are used; burr grinder and blade grinder. Home brewing people and most academic scientists usually use blade grinder for its low-cost, but this blade grinder cannot guarantee the evenness of the coffee powder and gives too much heat to the coffee to induce the loss of a lot of volatile compounds. However, there is no study for the composition changes by the grinders because it is known as not the critical point for the flavor changes than roasting, brewing or even the storages.

Most of the particle size research is conducted by the grinder sellers for showing their quality of evenness as a way of public relation. The particle sizes of the ground coffee beans are decided by the way of brewing. Typical particle sizes for their brewing methods are roughly from 250 µm, 500 µm, 1,000 µm and 1,500 µm for Espresso, Hand-brewing, Cold-brew and Frenchpress, respectively.

The main issues for the grinding step are size equalization and heat control. For both reasons, burr grinder is preferred over the blade grinder and the burr grinder is divided into two categories; flat burr and conical burr grinders. Flat burr has advantage in size equalization and conical burr has advantage in heat control. Therefore, for most café or companies, conical burr grinder is recommended because they grind plenty of coffee and it makes much heat, so heat control is more important for those. However, for home users, they can give enough time for cooling to the grinder, so they can accept temperature increasing for more even size.

1.2.4. Ground Coffee 1.2.4.1. Classifications The ground coffee should have the shortest period among the whole coffee process. By sharp growing of the surface area, the ground coffee becomes rancid quickly. Therefore, this step

does not have subordinate classifications, but to pass through the final cup of coffee by brewing.

1.2.4.2. Processes (Brewing / Extraction) The extracting process applied to ground coffee is called brewing. There are a lot of different coffee brewing techniques; hot and cold brewing, espresso brewing, machine-brewing, mocha pot, syphon, Turkish brewing and hand-brewing.

Machine-brewing and espresso brewing are now the most popular techniques for consuming coffee. Hand-brewing is one of the oldest techniques, but started to be re-illuminated by the increasing market of specialty coffee. Cold brew coffee is the most recent style and the unique coffee brewing style because it does not boil water to brew coffee. And it becomes popular with its unique flavor. However, since it is too new to be identified by scientist, there was no scientific publication for this remarkably growing brewing style.

Because changing brewing method is the only allowed variation for the general coffee consumers, a lot of different brewing methods were developed for specialty coffee society to control the flavor of coffee. On the other hand, because brewing cannot give enough energy for the components to have chemical reactions, coffee brewing methods were not attractive enough for the scientific researchers. However, with the coffee market grows, results started to come out that can fulfill both the curiosity for the specialty coffee society and scientific interpretations for the academia. 66

1.2.5. Extracted Coffee After brewing coffee, there seems no other process but to drink. However, there could be one more step to make the extracted coffee into solid powder. The product is called ‘instant coffee’ that the consumers can make a cup of coffee in a short time simply by pouring hot water because the solid is already extracted once. This method was first developed and patented in 1881, by Alphonse Allais, France, first started to sell with patent in 1890, by David Strang, New Zealand, first introduced to public in 1901, by Satori Kato, Japan, and first successfully commercialized in 1910, by George Washington, Belgium. Using instant coffee process or not, we can only categorize the coffee as hot or cold and pure or blended. Since there are too many factors for composition change exist in blending with milk, it is hard to discuss in scientifically. Therefore, most of the research on extracted coffee differences focused on taste

panel survey. Until now, it is accepted as truth that more bitter coffee harmonizes with milk better than less bitter coffee. Thus, acidic and sweet coffee is used for specialty coffee consumers and bitter and rich-body coffee is used for blended coffee consumers.

There is an emerging field of research about furan moiety. It is known that all the cooking food makes furan moiety and those compounds have potentially harmful for our health. By increasing coffee consumption, coffee is also regarded as an important source of furan family. Therefore, to reduce the furans would be an interesting issue for the further research.

1.3. Coffee Sensory Perception Coffee flavor perception is the combination of taste by the solid contents in the coffee liquid and smell by the volatile compounds. In general, when coffee consumers drink coffee, they can perceive the aroma, bitterness, body, sour, sweet and after taste properties. These components of coffee taste are provided through either by taste, smell or both. In this chapter, compounds comprising coffee taste and smell are reviewed and noble development of coffee taste classifications are embraced with Coffee Taster’s Flavor Wheel and World Coffee Research Sensory Lexicon.

1.3.1. Coffee Aroma The more coffee market advances, the more the coffee drinkers search for a better coffee taste. However, our taste perceptive system can only detect 5 different tastes with taste buds; sweet, sour, bitter, salty and umami. To achieve descriptions of various coffee tastes, the smell of the coffee is the basic essential, called coffee aroma.

For coffee aroma research, GC-MS is usually used for analyzing volatile compounds in coffee. In 2018, Nicola Caporaso et al. published the analysis of 50 aroma active volatile compounds in coffee samples by SPME (Solid phase microextraction)-GC-MS which is shown in Table 6.67

Table 6. Identification of volatile compounds in roasted arabica and robusta coffee samples analyzed by SPME-GC-MS at a single bean level.

The ultimate goal of coffee aroma research is to know what kind of coffee aroma is made by which coffee volatile compounds. This research is challenging because perception of flavor, the odor activity values and the perception threshold concentrations of the compounds are all different individually. Therefore, almost non-detectable composition change by scientific instruments can play a significant role in coffee taste. The differences of threshold concentrations of key food odorants are shown in Table 7.68

Table 7. Odorant qualities and threshold concentrations of selected key food odorants Key Food Odorants Odor quality Threshold Conc. ( µg/kg water) ethanol alcoholic 990,000 2-methyl-1-propanol malty 19,000 Acetic acid vinegar-like 5,600 1-hexanol green, grassy 590 (E)-2-hexenal green, apple-like 110 2-phenylethanol flowery, wine-like 18 (R)-limonene citrus-like 13 2-methoxy-4-vinylphenol smoky 5 3-hydroxy-4,5-dimethyl- seasoning-like 2 2(5H)-furanone butan-2,3-dione butter-like 1 3-methylbutanal malty 0.5 3-(methylthio)propanal cooked potato-like 0.4 (E)-2-hexenal green, grassy 0.1 2-acetyl-1-pyrroline popcorn-like 0.05 (E,E)-2,4-decadienal fatty, French fries-like 0.03 wine lactone coconut-like 0.02 (E)-β-damascenone cooked apple-like 0.01 (E,Z)-2,6-nonadienal cucumber-like 0.005 (Z)-1,5-octadien-3-one geranium-like 0.0003 1-p-menthene-8-thiol grapefruit-like 0.0002 2-methyl-3-furanthiol meaty, bouillon-like 0.00003

Unlike other food analytical studies, coffee aroma specifically depends most importantly on its process; roasting. The concentrations of the odorants in coffee have already been shown in Table 5, in the roasting chapter. Since the aroma of coffee changes significantly pursuant to the roasting degree, customized analytical equipment is needed. Therefore, the kinetic studies of volatile compounds produced during coffee roasting have been conducted using the roaster-connected Mass spectrometry customized only for the coffee research. 69

1.3.2. Coffee Taste Although coffee aroma plays more and more roles in specialty coffee trend than coffee taste, non-volatile coffee compounds take charge of the key perception in coffee – bitterness. The perception of bitterness is assumed to have developed for protecting human body from toxic compounds. Therefore, it was hypothesized that human perception has evolved regarding bitterness as an unpleasant taste. However, this hypothesis is not sufficiently supported 58

because only a little more than half of the bitter compounds are known as toxic to human. 70 Furthermore, coffee itself represents bitter beverage, and is known as a health beneficial beverage. The relationship between bitterness and toxicity needs to be further investigated.

In coffee, caffeine, chlorogenic acids and special sequences of amino acids with diketopiperazines are the main bitter taste compounds. Among them, the most popular bitter compound to coffee consumers is caffeine. However, only 1 % of the extracted coffee solid content is caffeine and its actual role on bitterness in a cup of coffee is only about 10 %. 33 This is easily identified that decaffeinated coffee, which has caffeine less than 3 % of normal coffee, is still bitter.

When the roasting degree is too high, less than 2 % of chlorogenic acids would be remained, and the bitterness would stem from the final products of Maillard reaction. 71 However, normal consuming coffee still conserve the main bitter taste compounds in coffee. Among the different bitterness raising compounds, chlorogenic acids play the largest part in coffee bitterness. About 80 % of bitterness in a cup of decaffeinated coffee originates from 10 major chlorogenic acids among the total coffee compounds.36

Special sequences of peptides and diketopiperazines are responsible for the remained bitterness of coffee. The bitterness of Leucine-containing peptides was selected to compare the differences of threshold concentration of bitterness by the sequences of amino acids. 72 Proline-based diketopiperazines are suspected to be important sources of coffee bitterness. 33

1.3.3. Tools for Coffee Sensory Analysis For the sensory analysis, panels are trained to perceive the coffee taste according to an official method. The most widely used training materials in coffee analysis are World Coffee Research Sensory Lexicon by Kansas University Coffee Research Center for the terminology and The Coffee Taster’s Flavor Wheel by SCA for actual training.

1.3.3.1. World Coffee Research Sensory Lexicon World Coffee Research started in 2012. Their research mainly focused on applying advanced agricultural science for coffee and now expended to all the area of coffee research including sensory analysis.

While working with coffee farms all around the world, they found the need for the unification on coffee taste expression, resulting in the first Sensory Lexicon published in 2016. The organization is a nonprofit organization. For the efficient terminology unification, the Sensory Lexicon is available free to all for downloading and printing of personal-use copies. The most recent publication version is 2.0, published in October, 2017. 73

The Sensory Lexicon includes attribute name, definition, references, intensity score and preparation instructions. By taste the samples prepared with the instruction, training panels can have reproducible taste terminology and unified intensity scores for different tastes in coffee.

1.3.3.2. The Coffee Taster’s Flavor Wheel The Coffee Taster’s Flavor Wheel was originally published in 1995 for training sensory panels by specialty coffee association of America (SCAA). It has been the industry standard for over two decades.

While World Coffee Research started to make its first Sensory Lexicon, SCAA was also one of the advisory group members. After WCR finished publishing the Sensory Lexicon successfully, it was essential to revise the SCAA Flavor Wheel to be compatible with the Lexicon as following the rationale behind the Lexicon publication. Therefore, the new Coffee Taster’s Flavor Wheel was revised in 2016.

The panel training can start from the center to outside in order to obtain easy relation between the attribute and their taste. It is also recommended to use only the core attributes (roasted, spices, nutty/cocoa, sweet, floral, fruity, sour/fermented, green/vegetable and others) when the panels are not well-trained. Panels can use higher attributes depending on their perception of the relation between the attribute and their taste. The colors of the attributes were selected with a careful consideration in order to match up the senses of sight with the senses of taste and smell. There are three different sized gaps between the attributes, and if there is a gap, which means the tasters thought of them as slightly less closely related.

References

1. Zhao Z, Moghadasian MH. Chemistry, natural sources, dietary intake and pharmacokinetic properties of ferulic acid: A review. Food Chem . 2008;109(4):691-702.

2. Ascherio A, Zhang SM, Hernán MA, et al. Prospective study of caffeine consumption and risk of parkinson's disease in men and women. Ann Neurol . 2001;50(1):56-63.

3. Cadden ISH, Partovi N, Yoshida EM. Review article: Possible beneficial effects of coffee on liver disease and function. Aliment Pharmacol Ther . 2007;26(1):1-7.

4. Fisher SM, McMurray RG, Berry M, Mar MH, Forsythe WA. Influence of caffeine on exercise performance in habitual caffeine users. Int J Sports Med . 1986;7(5):276-280.

5. Clifford MN, Johnston KL, Knight S, Kuhnert N. Hierarchical scheme for LC-MSn identification of chlorogenic acids. J Agric Food Chem . 2003;51(10):2900-2911.

6. Dulloo AG, Duret C, Rohrer D, et al. Efficacy of a green tea extract rich in catechin polyphenols and caffeine in increasing 24-h energy expenditure and fat oxidation in humans.

Am J Clin Nutr . 1999;70(6):1040-1045.

7. Reissig CJ, Strain EC, Griffiths RR. Caffeinated energy drinks-A growing problem. Drug

Alcohol Depend . 2009;99(1-3):1-10.

8. Macrae R, Robinson RK, Sadler MJ. Encyclopedia of food science, technology and nutrition. Academic Press; 1993.

9. Kuhnert N, Jaiswal R, Matei MF, Sovdat T, Deshpande S. How to distinguish between feruloyl quinic acids and isoferuloyl quinic acids by liquid chromatography/tandem mass spectrometry. Rapid Communications in Mass Spectrometry . 2010;24(11):1575-1582. 61

10. Kuhnert N, Jaiswal R, Eravuchira P, El-Abassy RM, von der Kammer B, Materny A.

Scope and limitations of principal component analysis of high resolution LC-TOF-MS data:

The analysis of the chlorogenic acid fraction in green coffee beans as a case study. Analytical

Methods . 2011;3(1):144-155.

11. Matei MF, Jaiswal R, Kuhnert N. Investigating the chemical changes of chlorogenic acids during coffee brewing: Conjugate addition of water to the olefinic moiety of chlorogenic acids and their quinides. J Agric Food Chem . 2012;60(49):12105-12115.

12. Sigma-Aldrich. Chlorogenic acid. http://www.sigmaaldrich.com/catalog/product/fluka/00500590. Updated 2015.

13. Wikipedia TFE. Chlorogenic acid. https://en.wikipedia.org/wiki/Chlorogenic_acid.

Updated 2015.

14. Yabuta G, Koizumi Y, Namiki K, Hida M, Namiki M. Structure of green pigment formed by the reaction of caffeic acid esters (or chlorogenic acid) with a primary amino compound.

Biosci Biotechnol Biochem . 2001;65(10):2121-2130.

15. Quideau S, Deffieux D, Douat-Casassus C, Pouysegu L. Plant polyphenols: Chemical properties, biological activities, and synthesis. Angewandte Chemie-International Edition .

2011;50(3):586-621.

16. Renaud S, Delorgeril M. Wine, alcohol, platelets, and the french paradox for coronary heart-disease. Lancet . 1992;339(8808):1523-1526.

17. Cho ES, Jang YJ, Hwang MK, Kang NJ, Lee KW, Lee HJ. Attenuation of oxidative neuronal cell death by coffee phenolic phytochemicals. Mutat Res -Fundam Mol Mech Mutag .

2009;661(1-2):18-24.

18. Haslam E. Practical polyphenolics - from structure to molecular recognition and physiological action. Cambridge University Press; 2005.

19. Ichihashi M, Ueda M, Budiyanto A, et al. UV-induced skin damage. Toxicology .

2003;189(1-2):21-39.

20. Valko M, Rhodes CJ, Moncol J, Izakovic M, Mazur M. Free radicals, metals and antioxidants in oxidative stress-induced cancer. Chem -Biol Interact . 2006;160(1):1-40.

21. Kuhnert N. One size does not fit all-bacterial cell death by antibiotics cannot be explained by the action of reactive oxygen species. Angewandte Chemie-International Edition .

2013;52(42):10946-10948.

22. Dawidowicz AL, Typek R. Transformation of 5-O-caffeoylquinic acid in blueberries during high-temperature processing. J Agric Food Chem . 2014;62(45):10889-10895.

23. Jaiswal R, Patras MA, Eravuchira PJ, Kuhnert N. Profile and characterization of the chlorogenic acids in green robusta coffee beans by LC-MSn: Identification of seven new classes of compounds. J Agric Food Chem . 2010;58(15):8722-8737.

24. Deshpande S, Jaiswal R, Matei MF, Kuhnert N. Investigation of acyl migration in mono- and dicaffeoylquinic acids under aqueous basic, aqueous acidic, and dry roasting conditions. J

Agric Food Chem . 2014;62(37):9160-9170.

25. Jaiswal R, Matei MF, Golon A, Witt M, Kuhnert N. Understanding the fate of chlorogenic acids in coffee roasting using mass spectrometry based targeted and non-targeted analytical strategies. Food & Function . 2012;3(9):976-984.

26. Iwai K, Kishimoto N, Kakino Y, Mochida K, Fujita T. In vitro antioxidative effects and tyrosinase inhibitory activities of seven hydroxycinnamoyl derivatives in green coffee beans.

J Agric Food Chem . 2004;52(15):4893-4898.

27. Kaiser N, Birkholz D, Colomban S, Navarini L, Engelhardt UH. A new method for the preparative isolation of chlorogenic acid lactones from coffee and model roasts of 5- caffeoylquinic acid. J Agric Food Chem . 2013;61(28):6937-6941.

28. Pascoa RNMJ, Magalhaes LM, Lopes JA. FT-NIR spectroscopy as a tool for valorization of spent coffee grounds: Application to assessment of antioxidant properties. Food Res Int .

2013;51(2):579-586.

29. Rubayiza A, Meurens M. Chemical discrimination of arabica and robusta coffees by fourier transform raman spectroscopy. J Agric Food Chem . 2005;53(12):4654-4659.

30. Deshpande S, El-Abassy RM, Jaiswal R, et al. Which spectroscopic technique allows the best differentiation of coffee varieties: Comparing principal component analysis using data derived from CD-, NMR- and IR-spectroscopies and LC-MS in the analysis of the chlorogenic acid fraction in green coffee beans. Analytical Methods . 2014;6(10):3268-3276.

31. Yeretzian C, Jordan A, Badoud R, Lindinger W. From the green bean to the cup of coffee:

Investigating coffee roasting by on-line monitoring of volatiles. Eur Food Res Technol .

2002;214(2):92-104.

32. Ginz M, Engelhardt UH. Identification of new diketopiperazines in roasted coffee. Eur

Food Res Technol . 2001;213(1):8-11.

33. Ginz M, Engelhardt UH. Identification of proline-based diketopiperazines in roasted coffee. J Agric Food Chem . 2000;48(8):3528-3532.

34. Frank O, Blumberg S, Kunert C, Zehentbauer G, Hofmann T. Structure determination and sensory analysis of bitter-tasting 4-vinylcatechol oligomers and their identification in roasted coffee by means of LC-MS/MS. J Agric Food Chem . 2007;55(5):1945-1954.

35. Frank O, Blumberg S, Kruempel G, Hofmann T. Structure determination of 3-O-caffeoyl- epi-gamma-quinide, an orphan bitter lactone in roasted coffee. J Agric Food Chem .

2008;56(20):9581-9585.

36. Frank O, Zehentbauer G, Hofmann T. Screening and identification of bitter compounds in roasted coffee brew by taste dilution analysis. Developments in Food Science . 2006;43:165-

168. doi: https://doi.org/10.1016/S0167-4501(06)80039-7.

37. Halliwell B, Gutteridge JMC. Oxygen free-radicals and iron in relation to biology and medicine - some problems and concepts. Arch Biochem Biophys . 1986;246(2):501-514.

38. Halliwell B, Clement MV, Long LH. Hydrogen peroxide in the human body. FEBS Lett .

2000;486(1):10-13.

39. Crozier A, Jaganath IB, Clifford MN. Dietary phenolics: Chemistry, bioavailability and effects on health. Nat Prod Rep . 2009;26(8):1001-1043.

40. Daglia M, Cuzzoni MT, Dacarro C. Antibacterial activity of coffee - relationship between biological-activity and chemical markers. J Agric Food Chem . 1994;42(10):2273-2277.

41. Kishimoto N, Kakino Y, Iwai K, Mochida K, Fujita T. In vitro antibacterial, antimutagenic and anti-influenza virus activity of caffeic acid phenethyl esters. Biocontrol

Sci . 2005;10(4):155-161.

42. Nakaso K, Ito S, Nakashima K. Caffeine activates the PI3K/akt pathway and prevents apoptotic cell death in a parkinson's disease model of SH-SY5Y cells. Neurosci Lett .

2008;432(2):146-150.

43. Hall S, Desbrow B, Anoopkumar-Dukie S, et al. A review of the bioactivity of coffee, caffeine and key coffee constituents on inflammatory responses linked to depression. Food

Res Int .

44. Runge FF. Kaffe. In: Phytochemical discoveries. ; 1820:144-159.

45. Yang A, Palmer AA, De Wit H. Genetics of caffeine consumption and responses to caffeine. Psychopharmacology . 2010;211(3):245-257.

46. Winston AP, Hardwick E, Jaberi N. Neuropsychiatric effects of caffeine. Adv Psychiatr

Treat . 2005;11(6):432-439.

47. Ludwig IA, Mena P, Calani L, et al. Variations in caffeine and chlorogenic acid contents of coffees: What are we drinking? Food Funct . 2014;5(8):1718-1726.

48. Consumer Reports magazine. The buzz on energy-drink caffeine. Consumer Reports magazine . December 2012.

49. Higdon JV, Frei B. Coffee and health: A review of recent human research. Crit Rev Food

Sci Nutr . 2006;46(2):101-123.

50. Johansson B, Halldner L, Dunwiddie TV, et al. Hyperalgesia, anxiety, and decreased hypoxic neuroprotection in mice lacking the adenosine A1receptor. Proc Natl Acad Sci U S A .

2001;98(16):9407-9412.

51. European Food Safety Authority (EFSA). Scientific opinion on the safety of caffeine.

EFSA Journal . 2015;13(5):4102.

52. Meyer JF, Jr., Roselius L, Wimmer K, inventors; Kaffee Handels Ag, assignee. Removing caffeine from coffee. 1908.

53. Saleeb FZ, Zeller BL, inventors; General Foods Corporation, assignee. Roasted coffee extract decaffeination method. 1985.

54. Noyes RM, Voeller RW, inventors; The Procter & Gamble Company, assignee. Coffee extraction process. 1982.

55. Zosel K, inventor; Studiengesellschaft Kohle m.b.H., assignee. Decaffeinizing coffee.

1971.

56. Fischer A, Kummer P, inventors; Coffex A.-G. S, assignee. Decaffeinization of raw coffee. 1980.

57. Kim IJ, inventorA Method For Processing Coffee Beans And The Coffee Beans

Therefrom. 2011.

58. Murthy PS, Manonmani HK. Physico-chemical, antioxidant and antimicrobial properties of indian monsooned coffee. Eur Food Res Technol . 2009;229(4):645-650.

59. Jumhawan U, Putri SP, Yusianto, Marwani E, Bamba T, Fukusaki E. Selection of discriminant markers for authentication of asian palm civet coffee (kopi luwak): A metabolomics approach. J Agric Food Chem . 2013;61(33):7994-8001.

60. Mayer F, Czerny M, Grosch W. Sensory study of the character impact aroma compounds of a coffee beverage. Eur Food Res Technol . 2000;211(4):272-276.

61. Semmelroch P, Grosch W. Studies on character impact odorants of coffee brews. J Agric

Food Chem . 1996;44(2):537-543.

62. Moon J, Yoo HS, Shibamoto T. Role of roasting conditions in the level of chlorogenic acid content in coffee beans: Correlation with coffee acidity. J Agric Food Chem .

2009;57(12):5365-5369.

63. Onakpoya I, Terry R, Ernst E. The use of green coffee extract as a weight loss supplement: A systematic review and meta-analysis of randomised clinical trials.

Gastroenterology research and practice . 2011;2011.

64. Federal Trade Commission. Green coffee bean manufacturer settles FTC charges of pushing its product based on results of “Seriously flawed” weight-loss study. . 2014.

65. Hemmerle H, Burger H-, Below P, et al. Chlorogenic acid and synthetic chlorogenic acid derivatives: Novel inhibitors of hepatic glucose-6-phosphate translocase. J Med Chem .

1997;40(2):137-145.

66. Gloess AN, Schönbächler B, Klopprogge B, et al. Comparison of nine common coffee extraction methods: Instrumental and sensory analysis. Eur Food Res Technol .

2013;236(4):607-627.

67. Caporaso N, Whitworth MB, Cui C, Fisk ID. Variability of single bean coffee volatile compounds of arabica and robusta roasted coffees analysed by SPME-GC-MS. Food Res Int .

2018;108:628-640.

68. Dunkel A, Steinhaus M, Kotthoff M, et al. Nature's chemical signatures in human olfaction: A foodborne perspective for future biotechnology. Angewandte Chemie-

International Edition . 2014;53(28):7124-7143.

69. Gloess AN, Yeretzian C, Knochenmuss R, Groessl M. On-line analysis of coffee roasting with ion mobility spectrometry–mass spectrometry (IMS–MS). Int J Mass Spectrom .

2018;424:49-57.

70. Nissim I, Dagan-Wiener A, Niv MY. The taste of toxicity: A quantitative analysis of bitter and toxic molecules. IUBMB Life . 2017;69(12):938-946.

71. Lund MN, Ray CA. Control of maillard reactions in foods: Strategies and chemical mechanisms. J Agric Food Chem . 2017;65(23):4537-4552.

72. Ishibashi N, Kanehisa H, Okai H, Fukui S. Bitterness of leucine-containing peptides.

Agric Biol Chem . 1987;51(9):2389-2394.

73. World Coffee Research. World coffee research sensory lexicon 2.0. . 2017.

2. Aim of the Research

The aim of this thesis is to investigate and ultimately control each steps of the coffee process to find how the chemical composition of the coffee bean changes using different process. The results are expected to tailor coffee for people with problems on caffeine or acid sensitivity.

Alternatively consumes that demands on coffee with increased levels of caffeine or chlorogenic acids can be satisfied.

The various aspects of this work include:

1) To find the effect of coffee grading and specialty coffee: Harvesting and grading step

Unlike coffee consuming trend changes to drink specialty coffee, high priced coffee

could not attract scientific interest because its rarity and high cost prevent from

changing the major coffee consuming style. Jamaica Blue Mountain coffee is one of

the most expensive and rare coffees in the world so did not attract scientific interest. In

this research, finding scientific reason why the coffee is special was conducted by

comparing different grades of Jamaica Blue Mountain coffees and other originated

normal price coffees.

2) To find the effect of roasting process and its alternative: Roasting step

Roasting has been an essential process for drinking coffee except consuming coffee

for medicinal use. However, the roasting process has a critical disadvantage that

degrades chlorogenic acids in coffee, which are regarded as the key compounds

responsible for health benefits in coffee. The noble method, which can substitute

roasting process using steaming and drying green coffee bean was developed in 2011.

However, the method was not analyzed in scientific methods but analyzed only with

sensory panels. This research is to obtain scientific data of the noble method and to 70

search pros and cons of the method compared to the normal roasting method.

3) To find the effect of extracting method: Brewing step

a. Extracting solvent

i. Comparison of water / methanol

ii. Comparison of different acidic water

b. Extracting method

Brewing step is the only adjustable stage for the coffee consumers among the whole

coffee processes. On the other hand, this step is the least interesting stage for chemists

because most of the brewing steps do not trigger chemical reactions. In this research,

the effect of different extracting solvent was compared with selected chlorogenic acids

in order to concatenate the academic coffee analysis result extracted by organic

solvents with practical water extracted coffee composition.

The kinetic study of coffee brewing was not conducted because it takes less than 20

seconds to brew an espresso coffee. Cold brew method has become popular since 2010

and it takes more than 12 hours so that it allows to study brewing kinetics. In this

research, the brewing kinetics using cold brew was studied focusing on the quantities

of chlorogenic acids and caffeine in coffee.

4) To find the applicability of coffee analyzing method to other beverages

All the analytical methods of different processes on coffee are not only useful for the

coffee analysis but also applicable to other food research. In this research, high priced,

specially parched, extracted with different solvents and kinetically obtained Korean

green tea samples were selected in order to prove the applicability of coffee analyzing

method to other food analysis.

3. Results 3.1. Research on Green Coffee Step

Analysis on different grades of highly-rated Jamaica Blue Mountain coffees

compared to easily available originated coffee beans

Seung-Hun Lee, Rakesh Jaiswal, Nikolai Kuhnert*

Department of Life Sciences and Chemistry, Jacobs University Bremen, 28759 Bremen,

Germany

*Correspondence to:

Prof. Dr. Nikolai Kuhnert,

Department of Life Sciences and Chemistry

Jacobs University Bremen gGmbH

Campus Ring 1

28759 Bremen, Germany

Tel: +49 421 200 3120

Fax: +49 421 200 3102

E-mail: [email protected]

Abstract

Origin and coffee grades are major factors determining the price of coffee beans.

Jamaica Blue Mountain coffee achieves among the highest prices in coffee trading due to its assumed high quality. In this contribution we show the first chemical composition analysis of Jamaica Blue Mountain using LC-MS methods. In particular differences in chlorogenic acid regioisomer distribution were observed for Jamaica Blue Mountain coffee if compared to reference coffees sheding light on the chemical basis of grading systems in coffee.

Keywords: Specialty coffee; Jamaica Blue Mountain coffee; Chlorogenic acid; tandem

Abbreviations Used

CG, caffeoyl glucose; (de), derivatives; QA, quinic acid; CQA, caffeoylquinic acid;

Hyd., hydrated; GCQA, glucose caffeoylquinic acid; CQAG, caffeoylquinic acid glucose; FQAG, feruloylquinic acid glucose; FQA, feruloylquinic acid; CQAL, caffeoylquinic acid lactone; CoQA, coumaroylquinic acid; CFQA, caffeoylferuloylquinic acid.

1. Introduction

Coffee is the third most consumed beverage after water and black tea in 21 st century.

According to the increasing economic importance of the coffee industry, numerous investigations with respect to the influence of the cup quality of coffee; species 1, origin 2, coffee production 3, roasting 4,5 and brewing 6 methods have been published.

With fast growing of coffee industry, decision standards of the price of coffee beans were set. Indonesian Luwak coffee became the most desired and expensive coffee in the world followed by the Black Ivory coffee, which appeared from Thailand around 2010.

However, these coffee products are additionally-artificially processed so that direct comparison to other normally processed coffees does not seem to be fair. The main decision standards of normally processed coffee price are species and origins. In general,

Arabica coffee is more expensive than Robusta coffee, and the price of same Arabica coffee differs by a factor of up to ten according to its origins. One of the most expensive coffees in the world is Jamaica Blue Mountain coffee which has a current retail price of

30 to 40 Euros per 250 g roasted, in Bremen, Germany, a factor of fifteen higher compared to standard supermarket products. Jamaica Blue Mountain coffee is praised for its rich aroma and long lasting after-taste.

The name Jamaica Blue Mountain comes from the grading system in Jamaica. Three different grading systems are commonly used in each country. One is grading by altitude of the coffee tree, another is by size of coffee bean and the other is by defected beans per bag. Jamaican Blue Mountain (JBM) grade is the grade name by altitude.

Jamaican coffee beans are divided Blue Mountain – High Mountain – Prime Washed –

Prime Berry by altitude of the coffee tree. Also, Jamaica Blue Mountain coffee is graded by size again, Blue Mountain Number 1 (Screen Size 17 ~ 18), Blue Mountain

Number 2 (Screen Size 16) and Blue Mountain Number 3 (Screen Size 15).

Furthermore, there is one more standard that if the beans are small and have some defected beans, the grade becomes Jamaica Blue Mountain Triage.

Although Jamaican coffee beans receive top evaluations from coffee societies following panel tasting, only limited scientific information rationalizing its high quality including environmental 7, isotope 8 and antioxidant assay 9 studies due to its rarity and high price is currently available. Therefore, this research is the first component analysis of Jamaican coffee beans by MS measurements including analysis of different JBM grades.

After the chlorogenic acids isomer study of Clifford et al. 10 , LC-MS n became one of the most important tools for analyzing different polyphenols in coffee so that a lot of coffee analysis researches with LC-MS n measurements are published. An extension of that,

HPLC-MS n is used for interpreting chemical structures in Jamaican coffee beans in this study.

2. Materials and Methods

The experimental methods of HPLC-LC/MS n are followed by previous publications of our group research 2,11,12 .

2.1. Chemicals and Coffee Beans

All the chemicals (Analytical grade) were purchased from Sigma-Aldrich (Germany).

Jamaica Blue Mountain Number 1, Jamaica Blue Mountain Number 2 and Jamaica Blue

Mountain Triage green coffee beans were provided by Youngki Moon from Jamaica,

Brazil, Colombia and Guatemala coffee beans were provided by Lloyd Caffee

(Fabrikenufer 115, Bremen, Germany).

2.2. Extract of Coffee Beans

All the green coffee beans were ground to fine powder, methanolic extracts were prepared by Soxhlet extraction using aqueous methanol (70 %, 170 mL) for 3 hours, using 5 g per each coffee beans. The extract was treated with Carrez reagents (1 mL of reagent I plus 1 mL of reagent II) to precipitate colloidal material and filtered through a

Whatman no. 1 filter paper and stored at -20 ℃ until required, thawed at room temperature, diluted to 60 mg / 10 mL, filtered through a membrane filter, and used for

LC/MS.

2.3. LC/MS n

The LC equipment (Agilent 1100 series, Karlsruhe, Germany) comprised a binary pump, an autosampler with a 100 µL loop, and a diode-array detector (DAD) with a light-pipe flow cell (recording at 320 and 254 nm and scanning from 200 to 600 nm). This was interfaced with an ion-trap mass spectrometer fitted with an electrospray ionization (ESI) source (Bruker Daltonics HCT ultra, Bremen, Germany) operating in full scan, auto

with a capillary temperature of 365 ℃, a drying gas flow rate of 10 L / min, and a nebulizer pressure of 10 psi.

2.4. HPLC

Separation was achieved on a 150 × 3 mm i.d. column containing diphenyl modified silica gel 5 µm, with a 5 mm × 3 mm i.d. guard column (Varian, Darmstadt, Germany).

Solvent A was water / formic acid (1000:0.005 v / v) and solvent B was methanol.

Solvents were delivered at a total flow rate of 500 µL / min. The gradient profile was from 10 to 80 % B linearly in 70 min followed by 10 min isocratic and a return to 10 %

B at 90 and 10 min isocratic to re-equilibrate.

3. Results and Discussion

3.1. Interpretations of chlorogenic acids

A total of three JBM green bean samples and three reference samples from Brazil,

Colombia and Guatemala were analyzed in this study. Extraction of green coffee beans and LC-MS analysis using reversed phase chromatography combined to negative ion mode MS monitoring using a quadrupole ion trap ESI-MS instrument were carried out as described previously.

As shown in Figure 1 , all six different coffees have qualitatively almost the same chlorogenic acid (CGA) composition. Three isomers of mono caffeoyl quinic acid and four isomers of dicaffeoyl quinic acids could be readily identified based on retention time and tandem MS provided in the literature 10,13 . Additional 40 further minor CGAs could be identified including feruloyl quinic acids (FQA), p-coumaroyl quinic acids ( p-

CoQA) and caffeoyl quinic acid lactones (CQAL). Total ion chromatograms (TICs) for six representative green bean coffee extracts are shown in Figure 1 . This similarity is not unexpected since all samples originated from the same botanical species; Coffea

Arabica . The common chlorogenic acids’ mass spectral data are shown in Table 1 with their chemical structures in Figure 2 .

Intens. x10 8 Jamaica Blue Mountain Number 1

0.8 0.6 0.4 0.2

Intens.0.0 x10 8 Jamaica Blue Mountain Number 2 1.0 0.8 0.6 0.4 0.2

Intens.0.0 x10 8 1.0 Jamaica Blue Mountain Triage 0.8 0.6 0.4 0.2

Intens.0.0 x10 8 0.8 Brazil Santos

0.6

0.4

0.2

Intens.0.0 x10 8 1.0 Colombia Supremo 0.8 0.6 0.4 0.2

Intens.0.0 x10 8 Guatemala Antigua 1.0 0.8 0.6 0.4 0.2 0.0 0 10 20 30 40 50 60 Time [min]

Figure 1. Total Ion Chromatogram of green coffee bean extracts in negative ion mode from six different coffees

Table 1. Tandem MS data of chlorogenic acids identified in coffee extracts of compounds in Jamaica Blue Mountain Number 1

MS 1 MS 2 MS 3

No. Compd. RT Base peak Secondary peak Base peak Secondary peak Parent ion m/z m/z int m/z int m/z m/z int m/z int m/z Int

1 CG (de) 3,7 341 178,6 160,6 34,6 112,8 12,91 160,6 142,6 65,16 89 62,38 112,8 47,43

2 CG (de) 3,7 473 340,8 130,8 21,81 452,8 3,43 178,6 160,6 74,78 281,6 19,68 118,8 17,57

3 CG 4 340,8 178,6 142,6 19,55 160,8 13,37 89 142,6 84,06 160,6 74,55 118,8 58,53

4 CG (de) 4,4 386,8 340,8 178,6 0,85 283 0,37 178,6 160,6 18,94 112,8 15,71 118,8 13,16

5 CG (de) 4,5 478 340,8 178,6 2,61 430 0,76 178,6 160,6 16,04 112,8 10,87 142,8 10,62

6 QA 5,9 190,6 85 93 66,84 126,8 65,86 59,2 -

7 QA (de) 6 405 190,6 386,8 10,38 274,6 6,58 126,8 172,6 100 85 57,26 110,8 46,01

8 CG (de) 6,4 341 178,6 322,8 74,34 142,6 24,76 160,6 89 95,71 124,6 63,19 100,8 58,9

9 CG (de) 6,9 457 340,8 178,6 5,72 438,8 4,14 178,6 142,8 18,5 160,6 17,03 118,8 9,83

10 CG (de) 11,5 487 383 441 33,78 341 8,94 340,8 322,8 18,71 178,4 2,72 112,8 2,11

11 Hyd. CQA 12,3 371 352,8 134,8 44,62 190,6 30,89 190,6 178,6 32,64 172,6 15,54 134,6 8,26

12 Hyd. CQA 12,8 371 352,8 134,8 43,79 190,6 41,44 190,6 178,6 46,51 134,8 11,51 172,6 10,66

13 Hyd. CQA 14,1 371 190,6 352,8 62,62 178,8 2,59 172,6 126,8 84,48 144,6 84,48 110,8 79,24

14 Hyd. CQA 15,2 371 352,8 172,6 64,8 190,6 28,45 172,6 190,6 63,82 178,6 54,62 134,6 7,8

15 Hyd. CQA 16,7 371 352,8 172,6 92,68 190,6 23,44 172,6 178,6 46,8 190,6 29,11 134,6 6,12

16 GCQA 17,9 515 178,6 340,8 29,78 352,8 28,56 134,8 -

17 CQAG 21,2 515 352,8 190,6 88,86 178,6 3,71 190,6 178,6 5,42 134,8 1,4 172,6 0,56

18 3-CQA 22,9 352,8 190,6 178,6 37,61 134,8 10,51 172,6 85 88,11 126,8 78,39 100,8 49,71

19 GFQA 25,6 529 190,6 367 83,05 172,6 58,72 93 175,8 32,47 85 19,56 143,4 4,06

20 CQAG 26,5 515 322,8 352,8 28,26 190,8 25,3 160,6 276,8 6,14 132,8 3,95 178,6 3,31

21 GCQA 26,7 515 352,8 322,8 98,86 340,8 93,59 190,6 172,6 93,73 178,6 43,99 134,8 8,13

22 MeOH-CQA 26,9 385 352,8 190,6 70,58 178,8 1,62 190,6 178,6 5,27 134,8 1,31 214,6 0,84

23 MeOH-CQA 27,4 385 352,8 190,6 78,68 351 32,4 190,6 214,6 21,29 178,6 5,1 172,6 3,14

24 5-CQA 27,8 352,8 190,6 214,6 7,42 178,6 2,86 172,6 85 85,57 110,8 59,36 126,6 59,27

25 3-CoQA 29,1 337 162,6 118,8 8,71 190,6 6,76 118,8 -

26 MeOH-CQA 31,3 385 352,8 172,6 17,28 190,6 9,16 172,6 178,6 61 190,6 12,9 214,6 7,79

27 3-CQAL 31,4 334,8 178,6 290,8 24,79 316,8 12,85 134,6 152,6 3,05 -

28 MeOH-CQA 31,5 385 352,8 172,6 23,21 178,6 8,15 172,6 178,6 62,52 190,6 18,08 134,8 8,75

29 3-FQA 32,2 366,8 192,6 193,6 8,84 133,8 7,78 133,6 148,6 26,47 190,6 3,12 116,8 2,58

30 4-CQA 33,5 352,8 172,6 178,6 54,46 190,6 21,98 93 110,8 88,73 154,6 32,97 71,2 26,15

31 4-CQAL 34,2 334,8 178,6 134,8 20,58 160,6 2,78 134,6 -

32 5-CoQA 35,3 336,8 190,6 162,6 4,36 172,6 1,95 126,6 85 100 172,6 66,2 93 56,48

33 5-FQA 37,1 366,8 190,6 192,6 3,16 172,8 2,36 126,8 172,6 54,36 85 52,09 108,8 29,29

34 4-CoQA 39,9 337 172,6 162,8 5,39 190,6 1,51 93 110,8 69,49 154,6 33,22 71,2 33,18

35 4-FQA 42,3 367 172,6 192,6 10,59 154,8 1,52 100,8 93 90,88 154,6 64,19 71,2 47,96

36 3,4,5-triCQA 43,6 677 515 334,8 15,6 353 13 352,8 178,6 20,66 340,8 16,86 334,8 12

37 3,5-diCQA 46,2 515 352,8 351 19,44 190,6 6,93 190,6 178,6 38,48 176,8 7,44 134,8 7,24

38 3,4-diCQA 48,5 515 352,8 334,8 24,61 172,6 19,04 172,6 178,6 71,95 190,6 53,6 134,8 9,15

39 4,5-diCQA 50 515 352,8 172,6 13,63 202,6 10,41 172,6 178,6 61,09 190,6 34,05 134,8 6,81

40 CFQA 51,4 529 366,8 352,8 81,75 351 17,16 192,6 190,6 46,8 172,6 16,43 133,6 9,79

41 CFQA 52,3 529 367 352,8 28,67 178,6 11,02 178,6 192,6 90,15 160,6 84,45 134,8 48,52

42 CQA (de) 53,5 499 352,8 318,8 74,15 336,8 36,67 172,6 178,6 45,97 190,6 17,94 134,8 7,37

43 p-CoQA (de) 54,7 499 336,8 360,8 59,9 318,8 32,49 172,6 335 83,63 162,8 25,32 190,8 5,85

44 CFQA 54,7 529 366,8 172,6 49,19 348,8 35,3 172,6 192,6 28,93 133,8 4,58 190,6 1,23

45 CFQA 56 529 367 352,8 90,57 350,8 40,31 172,6 192,6 52,04 190,6 35,43 133,8 5,95

46 p-CoQA (de) 61 513 336,8 348,8 30,8 172,8 26,62 172,6 162,6 17,14 190,6 1,79 294,8 1,52

47 FQA (de) 61,6 543 367 349 31,46 172,6 17,76 172,6 192,6 38,52 133,6 4 154,6 1,73

O OH OH HO O O OH O O OH O OGlc OH O GlcO OH O HO OH OH OH OH OH OH HO OH OH

HO OH CG QA Hyd. CQA GCQA O OH OH HO O O OH O OH O O O O O O OH OH O HO OH HO OH OH OGlc OH OGlc

HO OH CQAG FQAG MeOH-CQA OH O O OH O OH O HO O O O OH O HO O O O OH OH OH OH O O OH HO OH OH OH

HO O HO OH HO OH 3-CQA 4-FQA 4-CQAL 5-CoQA O O HO OH O OH OH O OH O O O HO HO OH O OH OH O OH O O O O O O OH O HO OH O OH HO OH HO OH O 3,4-diCQA 3-C-4-FQA

HO OH 3,4,5-triCQA

Figure 2. Chemical structures of chlorogenic acids identified in coffee extracts of compounds in Jamaica Blue Mountain Number 1

3.2. Similarity among the different grades in Jamaica Blue Mountain coffee beans.

As shown in Figure 1 , the three different graded Jamaica Blue Mountain coffee beans show qualitatively almost the same CGA composition. Therefore, without using any more sophisticated statistical tools and a larger number of samples, a comparison of among the Jamaica Blue Mountain coffee beans is meaningless and was not further pursued.

3.3. Different components between Jamaican and other originated beans

Although qualitatively the variety of chlorogenic acids are the same in all different originated coffee beans, the individual relative amounts of CGA regioisomers are different in Jamaican beans if compared to other reference beans.

In Figure 3 , extracted ion chromatograms (EICs) with m/z 353 and 515 corresponding to monocaffeoyl and dicaffeoyl quinic acid respectively are shown and the integrations values given in Table 2 , integration ratios of 3-CQA and 4-CQA against 5 CQA and

3,4-diCQA against the average of 5-CQA including diCQAs. The relative amount of 5-

CQA for mono caffeoyl derivatives and 3,4-diCQA for dicaffeoyl derivatives have been set to 100 % for ease of comparison.

Intens. x10 7 5-CQA 4-CQA 3,4-diCQA Jamaica Blue Mountain Number 1 4,5-diCQA 6 3-CQA 3,5-diCQA 4 2

Intens.0 x10 8 0.8 Jamaica Blue Mountain Number 2

0.6

0.4

0.2

Intens.0.0 x10 8 Jamaica Blue Mountain Triage 0.8

0.6

0.4

0.2

Intens.0.0 x10 7 Brazil Santos 6 4 2

Intens.0 x10 8 Colombia Supremo 0.8

0.6

0.4

0.2

Intens.0.0 x10 8 Guatemala Antigua 0.8

0.6

0.4

0.2

0.0 0 10 20 30 40 50 60 Time [min]

Figure 3. EIC of 353 (CQA) and 515 (diCQA) in six different coffees; 3-CQA, 5-CQA,

4-CQA, 3,5-diCQA, 3,4-diCQA and 4,5-diCQA peaks in order of retention time

Table 2. Relative amounts of monocaffeoyl quinic acids and dicaffeoyl quinic acids against 5-CQA and 3,4-diCQA set at 100 % for comparison obtained by integration of

EICs at m/z 353 and 515 respectively

5-CQA 3-CQA 4-CQA AVG 3,4-diCQA 3,5-diCQA 4,5-diCQA AVG

JBM Number 1 100 24,65 57,96 41,31 100 63,02 125,85 94,43

JBM Number 2 100 29,98 62,50 46,24 100 57,77 117,89 87,83

JBM Triage 100 29,32 66,88 48,10 100 63,61 129,23 96,42

Brazil Santos 100 18,15 42,31 30,23 100 158,92 124,74 141,83

Colombia Supremo 100 18,42 41,98 30,20 100 112,89 181,24 147,06

Guatemala Antigua 100 15,92 38,50 27,21 100 154,79 151,08 152,93

As shown in Table 2, the three Jamaican coffee beans contain considerably more 3- and

4-CQA with respect to total 5-CQA if compared to the reference coffee beans. All three averages of Jamaican 3- and 4-CQA contents are over 40 % of their 5-CQA contents whereas other originated beans are only around 30 %. Likewise, three Jamaican coffee beans have an increased 3,4-diCQA contents if compared to the relative amount of 5- substituted diCQAs. All three Jamaican average contents of 5-CQA containing diCQAs

(average of 3,5- and 4,5-diCQA) are less than their non-5-CQA containing diCQA (3,4- diCQA) while other originated coffee beans have around 150 %.

4. Conclusions

We analyzed Jamaica Blue Mountain coffee beans for the first time with HPLC-MS analysis and interpreted 47 compounds including chlorogenic acids. This interpretation informs us that the highly evaluated cup quality of coffee does not effect on the contents of coffee beans in the m/z range of 50 ~ 1500.

We found that different grades of Jamaican coffee beans are virtually identical with respect to their chemical composition.

There are differences between Jamaica Blue Mountain and other reference coffee beans especially with respect to their quantities of 3-CQA and 4-CQA. Levels of 3-CQA and

4-CQA are considerably higher in JBM coffee beans. Similarly 3,4-diCQA levels are considerably higher compared to their 5-substituted analogues. For the first time we could show that a coffee of a defined origin has a significantly altered profile of regioisomeric CGAs. We would like to suggest that this difference in regioisomer composition, observed here might be responsible for sensory differences of JBM if compared to lower valued coffee beans. 5-substituted CGAs are unable to form chlorogenic acid lactones in coffee roasting, a class of compounds that has been frequently linked to the sensory attributes of coffee 14,15 . Those could be connected to rich aroma of Jamaica Blue Mountain coffee to make this coffee as one of the cup quality coffee in the world.

Acknowledgements

Excellent technical support by Anja Müller, providing Jamaica coffee beans from

Youngki Moon and providing other originated coffee beans from Lloyd Caffee are acknowledged.

References

1. Lucon Wagemaker, T.A.; Limonta Carvalho, C.R.; Maia, N.B.; Baggio, S.R.; Guerreiro Filho, O. Sun protection factor, content and composition of lipid fraction of green coffee beans. Industrial Crops and Products 2011, 33, 469-473.

2. Kuhnert, N.; Jaiswal, R.; Eravuchira, P.; El-Abassy, R.M.; von der Kammer, B.; Materny, A. Scope and limitations of principal component analysis of high resolution LC-TOF-MS data: the analysis of the chlorogenic acid fraction in green coffee beans as a case study. Analytical Methods 2011, 3, 144-155.

3. Balyaya, K.J.; Clifford, M.N. Chlorogenic acids and caffeine contents of monosooned Indian Arabica and Robusta coffees compared with wet and dry processed coffees from the same geographic area. Colloq. Sci. Int. Cafe, [C. R. ] 1995, 16th, 316-325.

4. Moon, J.; Yoo, H.S.; Shibamoto, T. Role of Roasting Conditions in the Level of Chlorogenic Acid Content in Coffee Beans: Correlation with Coffee Acidity. J. Agric. Food Chem. 2009, 57, 5365-5369.

5. Moon, J.; Shibamoto, T. Role of Roasting Conditions in the Profile of Volatile Flavor Chemicals Formed from Coffee Beans. J. Agric. Food Chem. 2009, 57, 5823-5831.

6. Gross, G.; Jaccaud, E.; Huggett, A.C. Analysis of the content of the diterpenes cafestol and kahweol in coffee brews. Food and Chemical Toxicology 1997, 35, 547- 554.

7. Robinson, D.E.; Mansingh, A. Insecticide contamination of Jamaican environment. IV. Transport of residues from coffee plantations in the blue mountains to coastal waters in eastern Jamaica. Environ. Monit. Assess. 1999, 54, 125-141.

8. Weckerle, B.; Richling, E.; Heinrich, S.; Schreier, P. Origin assessment of green coffee (Coffea arabica) by multi-element stable isotope analysis of caffeine. Analytical and Bioanalytical Chemistry 2002, 374, 886-890.

9. Parras, P.; Martinez-Tome, M.; Jimenez, A.M.; Murcia, M.A. Antioxidant capacity of coffees of several origins brewed following three different procedures. Food Chem. 2007, 102, 582-592.

10. Clifford, M.N.; Johnston, K.L.; Knight, S.; Kuhnert, N. Hierarchical scheme for LC- MSn identification of chlorogenic acids. J. Agric. Food Chem. 2003, 51, 2900-2911.

11. Jaiswal, R.; Patras, M.A.; Eravuchira, P.J.; Kuhnert, N. Profile and Characterization of the Chlorogenic Acids in Green Robusta Coffee Beans by LC-MSn: Identification of Seven New Classes of Compounds. J. Agric. Food Chem. 2010, 58, 8722-8737.

12. Jaiswal, R.; Matei, M.F.; Golon, A.; Witt, M.; Kuhnert, N. Understanding the fate of chlorogenic acids in coffee roasting using mass spectrometry based targeted and non- targeted analytical strategies. Food & Function 2012, 3, 976-984.

13. Clifford, M.N.; Knight, S.; Kuhnert, N. Discriminating between the six isomers of dicaffeoylquinic acid by LC-MSn. J. Agric. Food Chem. 2005, 53, 3821-3832.

14. Frank, O.; Zehentbauer, G.; Hofmann, T. Bioresponse-guided decomposition of roast coffee beverage and identification of key bitter taste compounds. European Food Research and Technology 2006, 222, 492-508.

15. Dunkel, A.; Steinhaus, M.; Kotthoff, M.; Nowak, B.; Krautwurst, D.; Schieberle, P.; Hofmann, T. Nature's Chemical Signatures in Human Olfaction: A Foodborne Perspective for Future Biotechnology. Angewandte Chemie-International Edition 2014, 53, 7124-7143.

3.2. Research on Roasting Step

Analysis of Steamed (“Fermented”) Coffee Beans as a Roasted Coffee Substitute

Seung-Hun Lee 1, In Jae Kim 2, Sabur Badmos 1 and Nikolai Kuhnert* 1

Department of Life Sciences and Chemistry, Jacobs University Bremen, 28759 Bremen,

Germany

Coffeebio, A-B203, 144-3, Sangdaewon-dong, Jungwon-gu, Seongnam-si, Gyeonggi-do,

Korea

*Correspondence to:

Prof. Dr. Nikolai Kuhnert,

Department of Life Sciences and Chemistry

Jacobs University Bremen gGmbH

Campus Ring 1

28759 Bremen, Germany

Tel: +49 421 200 3120

Fax: +49 421 200 3102

E-mail: [email protected]

Abstract

In this contribution, steamed coffee beans (also referred to as fermented coffee) that have been introduced commercially as roasted coffee substitute were chemically profiled. Total polyphenol content and antioxidant capacities have been determined and compared to green and roasted coffee beans. Additionally, the chlorogenic acid profile has been determined using LC-MS methods and compared to green and roasted coffee beans. Steamed coffee shows a higher antioxidant capacity if compared to green and roasted coffee as well as an altered chlorogenic acid profile.

1. Introduction

From farm to table, diverse processes are used in each step of coffee production. As a first processing step the flesh is removed from the coffee cherry by dry, wet and semi- dry processes to obtain green coffee bean. 1,2 The term fermentation in coffee has already been widely used in the process of transforming coffee cherry to green coffee bean, and this refers normally to the action of microorganisms on green coffee beans for a period of three to seven days. 3 As a next step green coffee beans are roasted at temperatures between 160 and 250 °C for about 5 to 20 min. During roasting, the chemical composition of coffee changes dramatically with most of the desirable aroma compounds being produced. Following roasting, grinding of intact roasted beans is carried out using two different types of grinder; burr and blade grinders. After grinding coffee bean, Espresso machine, mocha pot, electric coffee maker or hand-brewing paper drip could be selected in personal preferences to achieve extraction of soluble coffee constituents under varying conditions. 4,5

To obtain the desirable organoleptic properties of coffee, roasting is considered as inevitable. However, at elevated temperatures chlorogenic acids (CGAs) are unavoidably decomposed to produce a series of chemically altered derivatives. This is accompanied by a reduction in total polyphenol content as measured by reduced antioxidant capacity assays. 6 On average, 50 % of the total CGAs are decomposed during roasting with extreme cases such as 98 % decomposition is already reported. 7

Since CGAs have been shown to be the main contributor to the beneficial health effects of the coffee against some human disorders such as Type-II diabetes, cardiovascular and

liver associated diseases, Parkinson syndrome etc., 8,9 coffee processing methods, which would retain or increase CGAs quantities, are highly desirable. To address this issue, a multinational company has introduced the Green Blend product containing 35 % of green coffee beans and hence delivering a beverage with high CGAs content. Increasing the quantity of green coffee beans to the product negatively affects the palatability and results in negative sensory reviews of the coffee beverage. Hence, alternative approaches towards producing a high CGA coffee that is acceptable to coffee consumers is of interest.

In 2011, a research on the use of steaming and drying of coffee was published as a patent which indicates that this special treatment of green coffee bean changed green flavor to fruit flavor, as well as the noble processed coffee bean extracted made it aroma richer than the reference-roasted coffee bean extracted by sensory evaluation. 10 This type of coffee is as well referred as “fermented coffee”. However, there were limitations in producing bitterness and burnt flavor which is usually regarded as unique properties of coffee taste.

Among the many methods on the coffee fermentation patent, steaming without active use of microorganisms gave a good condition with final fermented bean. The fermentation procedure used in this paper is this steaming and drying method only with the allowance of natural containing microorganisms to process the fermentation.

Therefore, the new word ‘steamentation’ is derived from this novel method which is in between steaming and fermentation. This steamentation process is a newly developed method of processing coffee which can possibly be a good substitute to the

monopolistic roasting method that has been employed for ages. This process involves starting from the green coffee bean which has already been fermented from the traditional system of coffee fermentation. It should as well be noted that “stomach friendly coffee” is frequently steam treated prior to roasting. 11,12

The aim of this study is to analyze antioxidants and to profile chlorogenic acids in steamented coffee bean using LC-MS techniques, purposely to compare it with the green and roasted coffee beans.

2. Experimental

The experimental methods of HPLC and LC-MS n were followed from previous publications. 13,14,15

2.1. Chemicals, Coffee Beans and Brewing Materials

All the chemicals (analytical grade) were purchased from Sigma-Aldrich (Germany).

Green, fermented and two differently roasted Brazil Arabica coffee beans were provided by Coffeebio, Gyeonggi-do, Korea.

2.2. Steamentation and Roasting Conditions of Coffee Beans

2.2.1. Steamentation

120 kg Brazil green coffee beans were steamed at 90 ℃ for 9 hr, dried at 50 ℃ for 15 hr and set at room temperature for 3 days. Those beans were steamed again at 72 ℃ for 7 hr and dried at 38 ℃ for 12 hr. In total, the steaming and drying processes were repeated

9 times which resulted into the steamented coffee bean as shown in Figure 1 .10

2.2.2. Roasting

The two differently roasted coffee beans are roasted using the same roaster (TOPER,

Turkey). Coffee beans were roasted in medium degree of roasting for handdrip roasted coffee bean and in between medium and city degree of roasting for espresso roasted coffee bean. Figure 1 depicts pictures of the respective coffee beans.

Figure 1. Pictures of green, steamented and two differently roasted coffee beans

2.3. Extract of Coffee Beans

All the green coffee beans were ground to fine powder, methanolic extracts were prepared by Soxhlet extraction using aqueous methanol (70 %, 170 mL) for 3 hr, using

5 g per different coffee beans. The extract was stored at -20 ℃ until required, thawed at room temperature, diluted to 5 % of the concentrated extract liquid, filtered through a membrane filter, and analyzed by LC/MS. The concentration of the extracted liquid were measured by evaporating solvent at 100 ℃, 30 min using aluminum boat, by measuring boat with solution before and after.

2.4. High Resolution LC-ESI-MS

The LC equipment (Agilent 1100 series, Karlsruhe, Germany) comprised a binary pump, an autosampler with a 100 µL loop, and a diode-array detector (DAD) with a light-pipe

flow cell (recording at 320 and 254 nm and scanning from 200 to 600 nm). This was interfaced with a MicroTOF Focus mass spectrometer (Bruker Daltonics, Bremen,

Germany) fitted with an electrospray ionization (ESI) source and internal calibration was achieved with 10 mL of 0.1 M sodium formate solution injected through a six port valve prior into each chromatographic run. Calibration was carried out using the enhanced quadratic mode. All experiments were carried out in negative ion mode.

2.5. LC-ESI-MS n

Daltonics HCT ultra, Bremen, Germany) operating in full scan, auto MS n mode (smart fragmentation) using a ramping of the collision energy. Maximum fragmentation amplitude was set to 1 V, starting at 20 % and ending at 200 %. MS operating conditions (negative mode) had been optimized using 5-caffeoylquinic acid with a capillary temperature of 365 ℃, a drying gas flow rate of 10 L / min, and a nebulizer pressure of 10 psi.

2.6. HPLC

Separation was achieved on a 150 × 3 mm i.d. column containing diphenyl 5 µm, with a

5 mm × 3 mm i.d. guard column (Varian, Darmstadt, Germany). Solvent A was water / formic acid (1000:0.005 v / v) and solvent B was methanol. Solvents were delivered at a

total flow rate of 500 µL / min. The gradient profile was from 10 to 80 % B linearly in

70 min followed by 10 min a return to 10 % and 10 min isocratic to re-equilibrate.

2.7. Antioxidant Assay

DPPH (2,2-diphenyl-1-picrylhydrazyl) and FRAP (Ferric ion reducing antioxidant power) assays were used to determine the antioxidant capacity of the coffee samples using 0.5 mg/mL garlic acid and 0.5 mg/mL Trolox (both dissolved in 70 % methanol) as standards. For the DPPH assay, 10 µL of the extracted coffee samples and standards were pipetted into 96-well plate with 200 µL of 0.5 mM DPPH solution added to each sample, mixed gently, incubated in the dark for 30 min and measured absorbance at 517 nm. For FRAP assay, freshly prepared FRAP reagent solution was used (10 mL of sodium acetate buffer mixed with 1 mL each of 10 mM solution of TPTZ (2,4,6-Tris(2- pyridyl)-s-triazine) in 40 mM HCl and 20 mM solution of Iron (III) chloride in water).

10 µL of each sample and standards were pipetted into 96-well plate, 200 µL of FRAP solution was added to each, mixed gently, incubated in the dark for 10 min at 37 ℃ and absorbance measured at 593 nm. 16-18

3. Results and Discussion

3.1. Changes in the Amount of Total Antioxidants as per Processes

The aim of this study is to compare the polyphenol content of steamented coffee with green and two differently roasted coffee beans. As roasted coffee beans, Brazilian

Arabica coffee beans roasted in medium degree of roasting (Hand drip roast) and in between medium and city degree of roasting (Espresso roast) were used as a comparison.

In analyzing the CGAs changes among the different processed coffees, 5-CQA was selected as an indicator compound because it is the most abundant chlorogenic acid in coffee. The amount of 5-CQA is quantified with HPLC-TOF-MS spectra with 280 nm

UV detector. The result is shown in Figure 2 .

3000

2500

2000

1500

5-CQA mg / L / mg 5-CQA 1000

500

0 Green Steamented Hand drip Espresso Roasted Roasted

Figure 2. Amount of 5-CQA in the four different processed coffee samples (mg / L)

Aqueous methanolic extracts of all the four samples were obtained and total antioxidant capacity was measured by using both DPPH and FRAP assays. All values are referenced to gallic acid and Trolox equivalents.

Previous studies have shown that the amount of antioxidant in coffee decreases by the increment of roasting degree.7 In the same vein, it is expected and proved with indicator quantification that the amount of total antioxidant would decrease in this order of green

– steamented – hand drip roasted – espresso roasted because as mentioned in 2.2.1., the steamentation process in this paper has maximum heating temperature at 90 ℃ which is between green (room temperature) and roasted coffee processes.

However, the antioxidant values shown in Figure 3 are different from initial expectation and indicator quantification data. The amount of antioxidant in steamented coffee beans was the highest among the four different processed coffee beans. This phenomenon could be explained with enhancing reactivity of degraded antioxidant compounds. In essence, the changes of antioxidant reactivity are higher than that of the reducing amount of antioxidant while antioxidant reducing is predominant factor in roasting process, even though the amount of total antioxidant is reduced by steamentation process.

3(a) DPPH g GAE / kg 70 g TE / kg

30 Amount in Sample in Amount 20

0 Green Steamented Hand drip Espresso Roasted Roasted

90 3(b) FRAP g GAE / kg 80 g TE / kg

30 Amount in Sample in Amount

0 Green Steamented Hand drip Espresso Roasted Roasted

Figure 3. Antioxidant assay by DPPH (a) and FRAP (b) for the green, steamented, handdrip roasted and espresso roasted coffee samples

3.2. Chlorogenic Acid Profiling

3.2.1. Analysis of high resolution MS data

To obtain information on the polyphenolics at a molecular level, all the four coffee samples were profiled by LC-MS techniques. High resolution MS data analysis, with a

LC-ESI-TOF-MS method, was used in obtaining high resolution m/z values and from this, the molecular formulae were determined. Table 1 shows all CGAs found in the

Brazil coffee beans listed by their retention times with their exact and theoretical m/z

100

values. Details of all the chemical structures and numberings in this paper are shown in the supporting information and follow our previous shorthand nomenclature on CGAs. 19

Table 1. High resolution mass (TOF-MS) data of chlorogenic acids

No. Name Mol. Formula Exp. m/z (M-H) Theor. m/z (M-H) Error (ppm)

1d 4-Caffeoylshikimic acid C16 H16 O8 335.0765 335.0772 2.2

2c 3-Caffeoylquinic acid lactone C16 H16 O8 335.0782 335.0772 -2.8

1e 5-Caffeoylshikimic acid C16 H16 O8 335.0770 335.0772 0.7

2d 4-Caffeoylquinic acid lactone C16 H16O8 335.0777 335.0772 -1.3

2a 1-Caffeoylquinic acid lactone C16 H16 O8 335.0775 335.0772 -0.9

3c 3-p-Coumaroylquinic acid C16 H18 O8 337.0929 337.0929 0

3e 5-p-Coumaroylquinic acid C16 H18 O8 337.0922 337.0929 2.1

3d 4-p-Coumaroylquinic acid C16 H18 O8 337.0931 337.0929 -0.7

4c 3-Caffeoylquinic acid C16 H18 O9 353.0888 353.0878 -2.7

4e 5-Caffeoylquinic acid C16 H18 O9 353.0901 353.0878 -6.4

4d 4-Caffeoylquinic acid C16 H18 O9 353.0885 353.0878 -2

5c 3-Feruloylquinic acid C17 H20 O9 367.1033 367.1035 0.5

5e 5-Feruloylquinic acid C17 H20 O9 367.1037 367.1035 -0.5

5d 4-Feruloylquinic acid C17 H20 O9 367.1041 367.1035 -1.8

6c Methyl-3-caffeoylquinate C17 H20 O9 367.1052 367.1035 -4.8

6d Methyl-4-caffeoylquinate C17 H20 O9 367.1051 367.1035 -4.3

7u Dimethoxycinnamoylquinic acid C18 H22 O9 381.1201 381.1191 -2.5

7u Dimethoxycinnamoylquinic acid C18 H22 O9 381.1205 381.1191 -3.5

8u Caffeoyl-coumaroylquinic acid C25 H24 O11 499.1245 499.1246 0.3

8u Caffeoyl-coumaroylquinic acid C25 H24 O11 499.1239 499.1246 1.5

9u Coumaroyl-feruloylquinic acid C26 H26 O11 513.1388 513.1402 2.9

10ce 3,5-diCaffeoylquinic acid C25 H24 O12 515.1197 515.1195 -0.3

10cd 3,4-diCaffeoylquinic acid C25 H24 O12 515.1199 515.1195 -0.8

10de 4,5-diCaffeoylquinic acid C25 H24 O12 515.1201 515.1195 -1.1

11u Caffeoyl-feruloylquinic acid C26 H26 O12 529.1379 529.1352 -5.1

11u Caffeoyl-feruloylquinic acid C26 H26 O12 529.1337 529.1352 2.7

11u Caffeoyl-feruloylquinic acid C26 H26 O12 529.1346 529.1352 1

11u Caffeoyl-feruloylquinic acid C26 H26 O12 529.1344 529.1352 1.5 Dimethoxycinnamoyl- 13u C H O 543.1532 543.1508 -4.4 feruloylquinic acid 27 28 12

3.2.2. Assignment of chlorogenic acids by tandem MS

Numerous tandem MS data allowing unambiguous assignment of CGAs to their respective regioisomeric level especially in the p-coumaroyl quinic acid derivatives;

101

caffeoyl quinic acids 20 , feruloylquinic acids 21 , and others 19 have been reported and were applied here.

Structures of CGAs in the four different coffee beans were assigned and are listed according to their retention time as shown in Table 2 along with MS data shown in

Figure 4. Table 2 shows the tandem MS data followed by compound numbering in

Supplementary table 1 and Supplementary figure 1 ~ 3 . And all of the four different processed coffee bean interpretations by fragmentation pattern are provided in

Supplementary table 2 ~ 5 . For compound abbreviations, we use the shorthand nomenclature introduced by Clifford MN.

102

Intens. 7 x10 4e Green Bean 3 3e 2 5c 4d 5e 1 4c 10ce 5d 10de 0 7 x10 4e Steamented Bean 1.25 2d 1.00 4ci 3ci 4d 10ce 10de 0.75 4ei 2c 4ei 4c 5d 0.50 3e 5e 0.25 5c 0.00 6 x10 Handdrip Roasted Bean 4e 1e 5 4ei 6c 2c 2d 5d 4 4ci 4ei 4d 12u 3 4c 5e 4ci 6d 2 5c 1 0 6 x10 Espresso Roasted Bean 6d 5d 6 4e 4d 5e 4ei 6c4ei 12u 4 4ci 2c 4c 1e 2d 2 5c

0 0 5 10 15 20 25 30 35 40 45 Time [min] Figure 4. Total ion chromatogram in negative ion mode of four different coffee beans

103

Table 2. Tandem MS data in negative ion mode of Chlorogenic acids from Brazil espresso roasted coffee bean extract

MS 1 MS 2 MS 3

No Compd. base peak Secondary peak base peak secondary peak Parent ion m/z m/z int m/z int m/z m/z int m/z int m/z int

1c 3-CSA 335.01 178.7 160.8 70 134.9 41 134.8 133.8 6 135.8 6 132.8 4

2c 3-CQAL 334.98 160.7 134.9 93 178.7 12 132.8 ------

1e 5-CSA 334.98 172.8 160.8 27 178.8 24 110.9 93 73 154.7 31 108.9 12

2d 4-CQAL 334.97 160.8 314.9 17 134.9 15 132.9 ------

2a 1-CQAL 334.97 160.8 172.7 83 315.9 12 132.8 ------

3e 5-pCoQA 336.96 190.8 134.9 7 160.7 6 84.9 172.8 57 87.2 53 - -

3d 4-pCoQA 336.97 172.7 162.8 6 173.7 6 110.9 93 48 98.9 24 - -

4c 3-CQA 352.46 190.8 178.8 44 176.8 34 85.1 172.7 71 126.9 66 108.9 34

4e 5-CQA 352.25 190.7 214.8 45 194.7 7 126.9 85.1 66 172.7 55 93.1 46

4d 4-CQA 352.47 172.8 178.8 51 190.8 24 93.1 110.9 69 154.7 16 59.6 13

5c 3-FQA 366.99 192.7 133.9 14 193.7 10 133.8 148.8 24 134.7 3 116.9 1

5e 5-FQA 367 190.8 191.7 7 192.8 6 126.9 172.8 74 85.2 71 110.9 57

5d 4-FQA 367.01 172.7 190.8 13 192.7 12 93 110.9 58 154.7 33 71.3 25

6c M-3-CQ 367 160.7 192.8 16 132.9 11 132.8 ------

6d M-4-CQ 366.99 160.8 134.9 36 192.7 10 132.9 ------

10ce 3,5-diCQA 515.02 353 351.1 33 353.9 9 190.7 178.7 53 176.8 15 134.9 8

10cd 3,4-diCQA 514.52 353 335 77 338.9 57 190.7 172.7 100 178.7 69 191.9 35

10de 4,5-diCQA 514.46 353 351 70 335 18 172.7 178.8 59 194.7 34 190.8 28

11ec 5-C-3-FQA 529 367 353 28 351 19 192.7 190.8 24 133.8 12 172.8 11

11u C-4-FQA 529.05 367 349 54 172.8 42 172.8 192.8 35 190.7 5 133.7 5

12u M-diCQA 529.03 367 160.8 20 368 19 160.8 134.8 26 132.9 14 192.7 11

3.3. Comparison of CGA profile in four coffee samples

When comparing the chromatogram of the four samples investigated, it becomes apparent that the number of peaks observed increases from green beans to steamented beans and to two differently roasted beans. This observation reflects that the degree of chemical transformation of CGAs undergo thermal treatment. Typical thermal reaction products such as CGA lactones and shikimic acid derivatives, both obtained through dehydration, transesterification products and epimers of quinic acid can be identified in

104

roasted bean samples. The most significant changes were observed when compared to steamented with roasted coffee beans for series of isobaric ions at m/z 367 and 335.

3.3.1. Differences in extracted ion chromatograms (EICs) at m/z 367

In EICs at m/z 367 ± 0.5, feruloylquinic acids and caffeoylquinic acid methyl esters corresponding to the molecular formula C 17 H20 O9 can be located. In all the four coffee bean samples three feruloylquinic acids were observed while signals corresponding to caffeoylquinic acid methyl ester were only observed in roasted coffee beans. With these

EICs, we observed that steamentation process decreases the relative intensity of FQAs, but does not yield methyl esters, which are only formed at elevated roasting temperatures. Figure 5 summarizes EICs at m/z 367 with their compound structures in

Supplementary figure 4 and their MS 2 spectra provided in Supplementary figure 5.

105

Intens. 05022015000002.D: EIC 367.00 -All MS 7 x10 5e 1.0 Green Bean

0.8

0.6

0.4 5d 0.2 5c

0.0 Intens. 05022015000003.D: EIC 367.00 -All MS x10 6 5e 3 Steamented Bean

2 5d 5c

0 Intens. 05022015000005.D: EIC 367.00 -All MS x10 6 5e 6d Handdrip Roasted Bean 2.0 5d 1.5 6c 6e 1.0 5c

0.5

0.0 Intens. 05022015000006.D: EIC 367.00 -All MS x10 6 5e Espresso Roasted Bean 2.5 6c 2.0 6d 1.5 5c 6e 5d 1.0

0.5

0.0 0 10 20 30 40 50 60 Time [min]

Figure 5. EICs at m/z 367 ± 0.5 (C 17 H20 O9, M-H) in negative ion mode of the four different processed coffee beans

3.3.2. Differences in EICs at m/z 335

In EICs at m/z 335 ± 0.5, caffeoylquinic acid lactones and caffeoylshikimic acids

(C 16 H16 O8, M-H) were observed while in the green bean spectrum, those signals at m/z

335 were absent.

In steamented, handdrip and espresso roasted coffee beans, 3- and 4-caffeoylquinic acid lactones were found. 1-CQL was observed only in roasted samples suggesting that acyl

106

migration does not take place during steamentation. Caffeoylshikimic acids were only observed in roasted coffee beans. With these observations, it could be concluded that steamentation process lead to elimination of water from CQAs to yield quinic acid lactones, while elimination of water from the cyclohexanol moiety to produce shikimic acid derivatives occurs only at higher temperatures as employed in roasting process.

Figure 6 shows the EICs of all the four samples with the chemical structures in

Supplementary figure 6; 4-CSA (1d), 3-CQAL (2c), 5-CSA (1e), 4-CQAL (2d) and 1-

CQAL (2a) respectively identified as species with a pseudo-molecular ion at m/z 335 by using MS 2 spectra shown in Supplementary figure 7. 22

107

Intens. 05022015000002.D: EIC 335.00 -All MS x10 5 1.0 Green Bean 0.8

0.6

0.4

0.2

0.0 Intens. 05022015000003.D: EIC 335.00 -All MS x10 5 2c Steamented Bean 4

3 2d

0 Intens. 05022015000005.D: EIC 335.00 -All MS x10 6 4 2c Handdrip Roasted Bean 3

2 2d 1e 1 1d 2a 0 Intens. 05022015000006.D: EIC 335.00 -All MS x10 6 2c 2.5 Espresso Roasted Bean 2.0 1e2d

1.5 1.0 1d 0.5 2a 0.0 0 10 20 30 40 50 60 Time [min]

Figure 6. EICs at m/z 335 ± 0.5 (C 16 H16 O8, M-H) in negative ion mode of the four different processed coffee beans

4. Conclusion

In conclusion, we analyzed Brazil Arabica coffee beans with LC-MS in four different treatment conditions; green coffee, steamented coffee, medium roasted coffee and high roasted coffee. We analyzed chlorogenic acids compounds with high resolution MS and interpreted polyphenolic compounds with tandem-MS spectra. In addition, antioxidant assays were processed with different standards and methods. Most importantly, we verified for the first time, the differences between steamentation and roasting processes.

Steamentation as a roasting substitute in coffee can make coffee drinkable as long as

108

chemical reactions are limited when compared with the settled roasting process. We suggest this research outcome would be useful in improving the trend on well-being style of coffee consumers for higher antioxidant containing coffee. Further research on various steamentation methods and combining it with roasting and steamenting processes would ultimately leads to production of higher antioxidant content and tastier coffee.

5. Acknowledgements

Excellent technical support by Anja Müller is acknowledged.

This Research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Conflicts of interest: none

References

1. Balyaya KJ, Clifford MN. Chlorogenic acids and caffeine contents of monosooned indian arabica and robusta coffees compared with wet and dry processed coffees from the same geographic area.

Colloq Sci Int Cafe, [C R ] . 1995;16th(Vol. 1, Seizieme Colloque Scientifique International sur le Cafe,

1995, Vol. 1):316-325.

2. Balyaya KJ, Clifford MN. Individual chlorogenic acids and caffeine contents in commercial grades of wet and dry processed indian green robusta coffee. Journal of Food Science and Technology-Mysore .

1995;32(2):104-108.

109

3. Amorim ACL, Hovell AMC, Pinto AC, et al. Green and roasted arabica coffees differentiated by ripeness, process and cup quality via electrospray ionization mass spectrometry fingerprinting. J Braz

Chem Soc . 2009;20(2):313-321.

4. Ludwig IA, Sanchez L, Caemmerer B, Kroh LW, Paz De Pena M, Cid C. Extraction of coffee antioxidants: Impact of brewing time and method. Food Res Int . 2012;48(1):57-64.

5. Sanchez-Gonzalez I, Jimenez-Escrig A, Saura-Calixto F. In vitro antioxidant activity of coffees brewed using different procedures (italian, espresso and filter). Food Chem . 2005;90(1-2):133-139.

6. Chandrasekara A, Naczk M, Shahidi F. Effect of processing on the antioxidant activity of millet grains.

Food Chem . 2012;133(1):1-9.

7. Moon J, Yoo HS, Shibamoto T. Role of roasting conditions in the level of chlorogenic acid content in coffee beans: Correlation with coffee acidity. J Agric Food Chem . 2009;57(12):5365-5369.

8. Cano-Marquina A, Tarín JJ, Cano A. The impact of coffee on health. Maturitas . 2013;75(1):7-21.

9. Shahidi F, Chandrasekara A. Hydroxycinnamates and their in vitro and in vivo antioxidant activities.

Phytochem Rev . 2010;9(1):147-170.

10. Kim IJ, inventorA Method For Processing Coffee Beans And The Coffee Beans Therefrom. 2011.

11. Rubach M, Lang R, Skupin C, Hofmann T, Somoza V. Activity-guided fractionation to characterize a coffee beverage that effectively down-regulates mechanisms of gastric acid secretion as compared to regular coffee. J Agric Food Chem . 2010;58(7):4153-4161.

12. Weiss C, Rubach M, Lang R, et al. Measurement of the intracellular ph in human stomach cells: A novel approach to evaluate the gastric acid secretory potential of coffee beverages. J Agric Food Chem .

2010;58(3):1976-1985.

110

13. Jaiswal R, Patras MA, Eravuchira PJ, Kuhnert N. Profile and characterization of the chlorogenic acids in green robusta coffee beans by LC-MSn: Identification of seven new classes of compounds. J Agric

Food Chem . 2010;58(15):8722-8737.

14. Jaiswal R, Matei MF, Golon A, Witt M, Kuhnert N. Understanding the fate of chlorogenic acids in coffee roasting using mass spectrometry based targeted and non-targeted analytical strategies. Food &

Function . 2012;3(9):976-984.

15. Kuhnert N, Jaiswal R, Eravuchira P, El-Abassy RM, von der Kammer B, Materny A. Scope and limitations of principal component analysis of high resolution LC-TOF-MS data: The analysis of the chlorogenic acid fraction in green coffee beans as a case study. Analytical Methods . 2011;3(1):144-155.

16. Huang DJ, Ou BX, Prior RL. The chemistry behind antioxidant capacity assays. J Agric Food Chem .

2005;53(6):1841-1856.

17. Zou Y, Chang SKC, Gu Y, Qian SY. Antioxidant activity and phenolic compositions of lentil (lens culinaris var. morton) extract and its fractions. J Agric Food Chem . 2011;59(6):2268-2276.

18. Borrelli RC, Visconti A, Mennella C, Anese M, Fogliano V. Chemical characterization and antioxidant properties of coffee melanoidins. J Agric Food Chem . 2002;50(22):6527-6533.

19. Jaiswal R, Kuhnert N. Hierarchical scheme for liquid chromatography/multi-stage spectrometric identification of 3,4,5-triacyl chlorogenic acids in green robusta coffee beans. Rapid Communications in

Mass Spectrometry . 2010;24(15):2283-2294.

20. Clifford MN, Johnston KL, Knight S, Kuhnert N. Hierarchical scheme for LC-MSn identification of chlorogenic acids. J Agric Food Chem . 2003;51(10):2900-2911.

21. Kuhnert N, Jaiswal R, Matei MF, Sovdat T, Deshpande S. How to distinguish between feruloyl quinic acids and isoferuloyl quinic acids by liquid chromatography/tandem mass spectrometry. Rapid

Communications in Mass Spectrometry . 2010;24(11):1575-1582.

111

22. Jaiswal R, Matei MF, Subedi P, Kuhnert N. Does roasted coffee contain chlorogenic acid lactones or/and cinnamoylshikimate esters? Food Res Int . 2014;61:214-227.

112

Supplementary table 1. List of chlorogenic acids

No Name Abb R1 R3 R4 R5 R1’ 1c 3-Caffeoylshikimic acid 3-CSA 2db C H H H 1d 4-Caffeoylshikimic acid 4-CSA 2db H C H H 1e 5-Caffeoylshikimic acid 5-CSA 2db H H C H 2a 1-Caffeoyl-1,5-quinide 1-CQL 5C H H 1 H 2c 3-Caffeoyl-1,5-quinide 3-CQL 5 C H 1 H 2d 4-Caffeoyl-1,5-quinide 4-CQL 5 H C 1 H 3c 3-p-Coumaroylquinic acid 3-CoQA H Co H H H 3d 4-p-Coumaroylquinic acid 4-CoQA H H Co H H 3e 5-p-Coumaroylquinic acid 5-CoQA H H H Co H 4c 3-Caffeoylquinic acid 3-CQA H C H H H 4d 4-Caffeoylquinic acid 4-CQA H H C H H 4e 5-Caffeoylquinic acid 5-CQA H H H C H 5c 3-Feruloylquinic acid 3-FQA H F H H H 5d 4-Feruloylquinic acid 4-FQA H H F H H 5e 5-Feruloylquinic acid 5-FQA H H H F H

6c 3-Caffeoylquinic acid methyl ester M-3-QA H C H H CH 3

6d 4-Caffeoylquinic acid methyl ester M-4-QA H H C H CH 3

6e 5-Caffeoylquinic acid methyl ester M-5-QA H H H C CH 3 7c 3-Dimethoxycinnamoylquinic acid 3-DQA H D H H H 7d 4-Dimethoxycinnamoylquinic acid 4-DQA H H D H H 7e 5-Dimethoxycinnamoylquinic acid 5-DQA H H H D H 8u Caffeoyl-coumaroylquinic acid CCoQA [C Co H H]* H 9u Coumaroyl-feruloylquinic acid CoFQA [Co F H H]* H 10cd 3,4-diCaffeoylquinic acid 3,4-diCQA H C C H H 10ce 3,5-diCaffeoylquinic acid 3,5-diCQA H C H C H 10de 4,5-diCaffeoylquinic acid 4,5-diCQA H H C C H 11u Caffeoyl-feruloylquinic acid CFQA [C F H H]* H

12u diCaffeoylquinic acid methyl ester M-diCQA [C C H H]* CH 3 13u Dimethoxycinnamoyl-feruloylquinic acid DFQA [D F H H]* H * ‘u’ is used for not confirmed isomeric positions in this paper, unconfirmed and ‘i’ is

used for stereoisomers. The stereoisomeric position numbering is shown in

Supplementary figure 1.

113

O R O 1 OR1'

R3O OR5 OR4

Supplementary figure1. Stereoisomeric positions of quinic acid

114

O OH O OH OH O HO O OH O O HO OH O OH O HO O OH OH OH OH HO OH 1c 1d 1e O HO OH O O O O HO O O OH OH HO O O O HO O OH OH OH O OH HO 2a 2c 2d O O HO OH OH OH O HO OH O O O O HO O OH HO O OH OH OH OH HO OH 3c 3d 3e O O HO OH OH OH O HO OH O OH O O O HO HO OH O OH HO O OH OH OH OH HO OH 4c 4d 4e O O HO OH OH OH O HO OH O O O O O O HO O O OH HO O OH OH OH OH HO OH 5c 5d 5e O O HO O O OH O HO O O OH O O O HO HO OH O OH HO O OH OH OH OH HO OH 6c 6d 6e O O HO OH HO OH OH OH O O O O O O O HO O O OH HO O O OH OH OH O O 7c 7d 7e

Supplementary figure2. Chemical structures of mono-chlorogenic acids

115

O HO OH O O HO OH O O OH HO OH

10ce O O HO OH HO OH O O HO OH O OH HO O O O HO O O OH 10cd 10de

HO OH HO OH

Supplementary figure3. Chemical structures of dichlorogenic acids

116

Supplementary table 2. Chlorogenic acids in Brazil green bean coffee interpretations MS 1 MS 2 MS 3

No Compd. base peak Secondary peak base peak secondary peak Parent ion m/z m/z int m/z int m/z m/z int m/z int m/z int

3e 5-pCoQA 337.03 190.8 191.7 8 162.8 5 126.8 93.1 34 172.8 29 85.2 23

4c 3-CQA 352.46 190.8 178.8 44 176.8 34 85.1 172.7 71 126.9 66 108.9 34

4d 4-CQA 352.98 172.8 178.8 46 190.8 17 93.1 110.9 93 154.8 47 71.3 26

4e 5-CQA 352.36 190.8 214.8 30 176.8 5 126.9 85.2 100 172.8 67 93.1 49

5c 3-FQA 366.99 192.7 133.9 14 193.7 10 133.8 148.8 24 134.7 3 116.9 1

5d 4-FQA 367.01 172.7 190.8 13 192.7 12 93 110.9 58 154.7 33 71.3 25

5e 5-FQA 367.02 190.8 191.7 7 192.8 6 126.8 85.2 61 172.8 60 93.1 49

10ce 3,5-diCQA 515.02 353 351.1 33 353.9 9 190.7 178.7 53 176.8 15 134.9 8

10de 4,5-diCQA 514.46 353 351 70 335 18 172.7 178.8 59 194.7 34 190.8 28

11 5-C-3-FQA 529 367 353 28 351 19 192.7 190.8 24 133.8 12 172.8 11

11 C-4-FQA 529.05 367 349 54 172.8 42 172.8 192.8 35 190.7 5 133.7 5

117

Supplementary table 3. Chlorogenic acids in Brazil fermented bean coffee interpretations MS 1 MS 2 MS 3

No Compd. base peak Secondary peak base peak secondary peak Parent ion m/z m/z int m/z int m/z m/z int m/z int m/z int

2c 3-CQAL 334.99 160.7 134.9 82 178.7 14 132.8 ------

2d 4-CQAL 334.97 160.8 314.9 17 134.9 15 132.9 ------

3c 3-pCoQA 337.03 162.8 163.8 11 190.8 9 118.9 ------

3d 4-pCoQA 337.03 172.8 173.7 9 162.8 7 93 110.9 62 - - - -

3e 5-pCoQA 337 190.8 162.8 7 191.7 7 93.1 126.9 50 87.1 30 85.1 29

4c 3-CQA 352.97 190.8 178.8 50 176.8 32 126.9 85.2 70 93.1 58 172.8 54

4ci 3-CQA de 352.96 190.8 178.8 59 176.8 17 ------

4ci 3-CQA de 352.97 190.8 178.8 57 176.8 21 93.1 110.9 65 126.8 51 172.7 49

4d 4-CQA 353 172.8 178.8 43 190.8 15 93.1 110.9 47 71.4 14 154.8 12

4e 5-CQA 352.25 190.7 214.8 45 194.7 7 126.9 85.1 66 172.7 55 93.1 46

4ei 5-CQA de 352.96 190.8 214.8 51 176.8 15 86.2 ------

4ei 5-CQA de 352.98 190.8 178.8 10 306.9 8 85.1 126.8 40 110.9 30 - -

5c 3-FQA 367.02 192.8 133.9 16 193.7 10 133.8 148.8 22 134.8 3 149.8 3

5d 4-FQA 367.03 172.7 192.8 14 173.6 9 93 110.9 30 71.3 18 154.7 11

5di 4-FQA de 366.99 172.8 347.9 17 192.8 14 71.3 136.8 36 154.8 30 93.1 23

5e 5-FQA 367 190.8 191.7 7 192.8 6 126.9 172.8 74 85.2 71 110.9 57

10ce 3,5-diCQA 515.04 352.9 351.1 35 190.8 11 190.7 178.8 53 134.9 14 132.9 12

10de 4,5-diCQA 514.46 353 351 77 172.8 19 172.7 178.7 67 176.8 31 190.8 24

118

Supplementary table 4. Chlorogenic acids in Brazil handdrip roasted bean coffee interpretation MS 1 MS 2 MS 3

No Compd. base peak Secondary peak base peak secondary peak Parent ion m/z m/z int m/z int m/z m/z int m/z int m/z int

1c 3-CSA 335.03 178.8 160.8 66 134.9 41 134.8 132.8 6 - - - -

1e 5-CSA 334.99 172.7 160.8 29 178.7 21 93.1 110.9 93 154.8 62 71.3 34

2a 1-CQAL 334.97 160.8 172.7 83 315.9 12 132.8 ------

2c 3-CQAL 334.98 160.7 134.9 93 178.7 12 132.8 ------

2d 4-CQAL 334.99 160.8 134.8 18 178.7 17 132.8 ------

3e 5-pCoQA 337.02 190.8 160.8 9 134.8 9 126.8 172.8 84 85.2 66 110.9 49

4c 3-CQA 352.98 190.8 178.8 44 176.7 35 126.8 93.1 60 85.1 52 170.7 43

4ci 3-CQA de 352.99 190.8 178.8 59 134.8 20 71.4 ------

4ci 3-CQA de 353 178.8 172.8 30 134.9 23 134.8 132.8 83 133.8 6 135.8 4

4ci 3-CQA de 353.02 190.8 178.8 73 176.8 22 93.1 ------

4d 4-CQA 352.98 172.8 178.8 46 190.8 18 93.1 110.9 35 154.7 20 71.3 16

4di 4-CQA de 352.95 172.8 178.8 39 190.8 36 154.8 93.1 97 - - - -

4e 5-CQA 352.42 190.8 214.8 38 176.8 7 85.2 126.9 98 93.1 73 172.7 52

4ei 5-CQA de 352.46 190.7 214.8 20 172.8 19 126.9 172.7 80 85.2 50 170.8 46

4ei 5-CQA de 353 190.8 214.8 34 176.7 6 172.8 85.1 60 93.1 58 170.8 57

5c 3-FQA 367.01 192.7 133.9 12 193.7 8 133.8 148.8 21 116.9 6 134.8 4

5d 4-FQA 367.04 172.7 192.8 12 173.7 7 93.1 111 39 154.8 39 71.4 25

5e 5-FQA 367.01 190.8 191.7 8 192.8 6 126.9 85.1 94 93 52 172.7 46

6c M-3-CQ 367 160.7 192.8 16 132.9 11 132.8 ------

6d M-4-CQ 366.99 160.8 134.9 36 192.7 10 132.9 ------

12 M-diCQA 529.03 367 160.8 20 368 19 160.8 134.8 26 132.9 14 192.7 11

119

Supplementary table 5. Chlorogenic acids in Brazil espresso roasted bean coffee interpretation MS 1 MS 2 MS 3

No Compd. base peak Secondary peak base peak secondary peak Parent ion m/z m/z int m/z int m/z m/z int m/z int m/z int

1c 3-CSA 335.01 178.7 160.8 70 134.9 41 134.8 133.8 6 135.8 6 132.8 4

1e 5-CSA 334.98 172.8 160.8 27 178.8 24 110.9 93 73 154.7 31 108.9 12

2c 3-CQAL 334.99 160.8 134.9 78 161.8 10 132.8 ------

2d 4-CQAL 334.98 160.7 178.8 29 134.9 24 132.8 ------

3d 4-pCoQA 336.97 172.7 162.8 6 173.7 6 110.9 93 48 98.9 24 - -

3e 5-pCoQA 336.96 190.8 134.9 7 160.7 6 84.9 172.8 57 87.2 53 - -

4c 3-CQA 352.97 190.8 178.8 46 176.8 35 85.2 172.8 69 110.9 52 126.9 47

4ci 3-CQA de 352.99 190.8 178.7 75 134.8 23 170.8 ------

4ci 3-CQA de 352.98 178.7 172.8 28 176.8 22 134.8 132.8 94 133.8 9 135.8 7

4d 4-CQA 352.47 172.8 178.8 51 190.8 24 93.1 110.9 69 154.7 16 59.6 13

4di 4-CQA de 352.97 172.8 178.8 44 190.8 42 110.9 93 89 154.7 36 - -

4e 5-CQA 352.47 190.7 214.8 40 176.8 7 85.1 172.8 69 126.8 62 93 59

4ei 5-CQA de 352.41 190.7 214.8 20 172.8 19 126.8 172.8 42 108.9 32 85.1 28

4ei 5-CQA de 352.98 190.8 214.8 30 172.8 7 110.9 172.8 93 126.9 86 93 79

5c 3-FQA 367.04 192.8 133.9 13 193.7 9 133.8 148.8 16 134.7 5 116.9 2

5d 4-FQA 367.02 172.8 192.7 10 173.7 7 93.1 154.8 40 110.9 33 71.3 24

5e 5-FQA 367.01 190.7 191.7 6 192.8 6 126.8 172.8 90 85.2 78 93.1 67

6c M-3-CQ 367.02 160.7 192.7 16 134.9 12 132.9 ------

6d M-4-CQ 367.03 160.8 134.9 38 132.9 10 132.8 ------

10cd 3,4-diCQA 514.52 353 335 77 338.9 57 190.7 172.7 100 178.7 69 191.9 35

10ce 3,5-diCQA 515.04 353 351 30 353.9 11 190.8 178.7 42 176.7 17 134.8 10

12 M-diCQA 529.04 367 367.9 18 160.8 17 160.7 134.8 15 192.8 9 132.8 8

120

O O HO OH O HO HO OH OH O HO O O O O O O OH HO O HO OH OH O OH HO OH 5c 5d 5e

O O HO O HO HO HO O OH O HO O O HO O OH O OH HO O HO O OH O OH HO OH 6c 6d 6e

Supplementary figure 4. Chemical structures of the peaks in EICs at m/z 367

121

Supplementary figure 5. MS 2 spectra of espresso roasted coffee bean from EIC at m/z

367

122

O OH O OH OH O O HO OH O O HO OH O OH O HO O OH OH OH OH HO OH 1c 1d 1e O HO OH O O O HO O O O OH OH HO O O O HO O OH OH OH O OH HO 2a 2c 2d

Supplementary figure 6. Chemical structures of the peaks in EICs at m/z 335

123

Supplementary figure 7. MS 2 spectra of handdrip roasted coffee bean from EIC at m/z

335

124

3.3. Research on Brewing Step 3.3.1. Research on Brewing Step with Different Solvents

Characterization of the chlorogenic acids composition with different extracting solvents

Seung-Hun Lee, Nikolai Kuhnert*

Department of Life Sciences and Chemistry, Jacobs University Bremen, 28759 Bremen,

Germany

*Correspondence to:

Prof. Dr. Nikolai Kuhnert,

Department of Life Sciences and Chemistry

Jacobs University Bremen gGmbH

Campus Ring 1

28759 Bremen, Germany

Tel: +49 421 200 3120

Fax: +49 421 200 3102

E-mail: [email protected]

125

Abstract

Methanol was used as an extracting solvent for analyzing coffee due to its low viscosity, price and customs of food analysis. The real consuming compounds are analyzed in addition to a large amount of scientific coffee results on account of knowing the different solubility between water and methanol extracted coffee. Nevertheless, scientific analysis of coffee is still processing methanol extraction. In this report, we interpreted various chlorogenic acids and confirmed chemical reactions during roasting time with known methods by analyzing tandem MS spectra of both water and methanol extracts of green and roasted Arabica Colombia Supremo. Furthermore, we identified the differences of relative composition between water and methanol extracts, focused on seven major compounds’ relative intensities against 5-CQA in extracted ion chromatograms.

Keywords: Tandem MS; Chlorogenic acid; Coffee extracting solvent

126

Abbreviations used

QA, quinic acid; CA, caffeic acid; CQA, caffeoylquinic acid; FQA, feruloylquinic acid;

CQAL, caffeoylquinic acid lactone; CoQA, coumaroylquinic acid; M-CQ, methyl- caffeoylquinate; CFQA, caffeoyl-feruloylquinic acid.

127

1. Introduction

Coffee is one of the most consumed beverages in the world. In order to get the best way for the taste of coffee, lots of analysis on all the coffee processes have been studied; species 1, coffee production 2, roasting 3-5 and brewing 6. Furthermore, not only analyzing, but also in-depth chemical studies into the chemical reactions during roasting process are proceeding briskly 7.

A lot of analysis tools are imported to analyze coffee components; HPLC 8,9 , GC 4,10 , LC-

MS 11,12, FT-IR 13,14 , UV-Vis Spectrometry 15 , NMR 16 , and so on. After the study of chlorogenic acids by tandem-MS technique 17 , LC-MS n became the mainstream for interpreting isomeric p-coumaric acid derivatives. This research also proceeded with tandem-MS for verifying peaks of isomeric polyphenols.

Chemical analysis of food has been processed with methanolic extracts. Most of the published coffee researches also used methanol as an extracting solvent in order to extract as much components as possible. However, especially for the coffee and tea, not like other food, we rarely eat them, but drink only water extracted part. Although practical research are done with water extract by governmental organizations, such as

Food and Drug Administration, scientific views are still focused on methanolic extracts, with regarding coffee and tea as other food.

This trend caused unnecessary double research. Although there already are a lot of well- interpreted scientific results done in laboratories, additional experiments were needed when people wanted to know the data of real-consumed coffee. Thus, this research

128

ascertain the effect of extracting solvent, identify the difference between methanol and water extracted compounds for opening the possibility to apply existing methanolic scientific data to watery real-consuming data, and help people plan their further research to be easily applicable to the real consumption.

2. Experimental

The experimental methods of HPLC and LC/MS n are followed by previous publications of our group research 1,7,11 .

2.1. Chemicals and Coffee Beans

All the chemicals (Analytical grade) were purchased from Sigma-Aldrich (Germany).

Green Colombia Supremo Arabica coffee bean was purchased from Lloyd Caffee in

Bremen, Germany.

2.2. Roasting

Colombia Supremo Arabica green coffee bean (100 g) was roasted with a Gene Café

Coffee Roaster 101A (Genesis Co. Kyungki, Korea), set temperature 240 ℃ for 15 min.

2.3. Extract of Coffee Beans

Colombia Supremo Arabica coffee beans (5 g per each sample, Green bean and 240 ℃,

15 min roasting) were ground to fine powder; methanolic extracts were prepared by soxhlet extraction using aqueous methanol (70 %, 170 mL) for 3 hours and water extracts were prepared by stirring in 100 ℃ water (200 mL) for 1 hour in 250 mL beaker. The extract was treated with Carrez reagents (1 mL of reagent A plus 1 mL of

129

reagent B) to precipitate colloidal material and filtered through a Whatman no. 1 filter paper and stored at -20 ℃ until required, thawed at room temperature, diluted to 60 mg /

10 mL, filtered through a membrane filter, and used for LC/MS.

2.4. LC/MS n

MS n mode (smart fragmentation) using a ramping of the collision energy. Maximum fragmentation amplitude was set to 1 V, starting at 20 % and ending at 200 %. MS operating conditions (negative mode) had been optimized using 5-caffeoylquinic acid with a capillary temperature of 365 ℃, a drying gas flow rate of 10 L / min, and a nebulizer pressure of 10 psi.

2.5. HPLC

Separation was achieved on a 150 × 3 mm i.d. column containing diphenyl 5 µm, with a

5 mm × 3 mm i.d. guard column (Varian, Darmstadt, Germany). Solvent A was water / formic acid (1000:0.005 v / v) and solvent B was methanol. Solvents were delivered at a total flow rate of 500 µL / min. The gradient profile was from 10 to 70 % B linearly in

60 min followed by 10 min isocratic and a return to 10 % B at 90 and 10 min isocratic to re-equilibrate.

130

3. Results and Discussion

3.1. Interpretations of chlorogenic acids

Interpretation methods of chlorogenic acids by LC/MS n have been already studied especially in p-coumaroyl quinic acid derivatives; caffeoyl quinic acids 17 ,

Feruloylquinic acids 18 , and others 19 .

Compounds were interpreted in four different spectra as shown in Table 1 ~ 4 with their

Total MS data in Figure 1 .

Figure 1. Total MS spectra of coffee beans in four different conditions

131

Table 1. Colombia Supremo Green Bean Water Extraction Interpretations MS 1 MS 2 MS 3

No Compd. base peak Secondary peak base peak secondary peak Parent ion m/z m/z int m/z int m/z m/z int m/z int m/z int

1 Sucrose 341.0 178.6 160.6 22 112.8 15 89 160.6 85 142.6 77 118.8 62

2 QA 190.6 126.8 85.0 99 172.6 74 108.8 98.8 16

3 QA ep. 190.6 110.6 172.6 37 67.4 -

4 3-CQA 352.8 190.6 178.6 41 134.8 9 126.6 172.6 69 85.0 65 110.8 31

5 5-CQA 352.8 190.6 - 126.8 172.6 100 85.0 62 93.0 38

6 3-FQA 366.8 192.6 133.8 11 133.6 148.6 34

7 4-CQA 352.8 172.6 178.6 57 190.6 15 92.8 110.8 87 154.6 40 71.2 24

8 5-CoQA 336.8 190.6 - 126.6 85.0 78 172.6 58 110.8 42

9 5-FQA 366.8 190.6 - 126.6 172.6 86 85.0 78 93.0 49

10 4-CoQA 336.8 172.6 162.6 7 190.6 4 110.8 71.2 27 154.6 26 81.0 21

11 4-FQA 366.8 172.6 192.6 13 92.8 110.8 89 71.2 34 154.6 28

12 3,5-diCQA 514.8 352.8 190.8 8 190.4 178.8 40 134.8 7

13 3,4-diCQA 514.8 352.8 172.6 23 334.8 19 172.6 178.6 65 190.6 63 134.8 8

14 4,5-diCQA 514.8 352.8 172.8 16 202.6 16 172.6 178.6 62 190.6 28 134.8 10

Table 2. Colombia Supremo Green Bean Methanol Extraction Interpretations MS 1 MS 2 MS 3

No Compd. base peak Secondary peak base peak secondary peak Parent ion m/z m/z int m/z int m/z m/z int m/z int m/z int

1 Sucrose 376.8 340.8 214.6 15 178.6 160.6 19 112.8 16 142.6 16

2 QA 190.6 126.6 172.6 100 85.0 99 108.8 98.8 16 83.0 14

3 QA ep. 190.6 110.8 172.6 30 67.2 -

4 3-CQA 352.8 190.6 178.6 45 134.8 9 172.6 126.6 92 85.0 78 110.8 55

5 5-CQA 352.8 190.6 - 126.6 172.6 71 85.0 64 93.0 43

6 3-FQA 366.8 192.6 133.8 11 133.6 148.6 28

7 4-CQA 352.8 172.6 178.6 59 190.6 17 93.0 110.8 94 154.6 56 71.2 45

8 5-CoQA 336.8 190.6 162.8 5 126.6 172.6 81 85.0 63 93.0 61

9 5-FQA 366.8 190.6 - 126.8 172.6 100 85.0 73 93.0 50

10 4-CoQA 336.8 172.6 162.6 6 93.0 71.2 71 110.8 71 154.6 24

11 4-FQA 366.8 172.6 192.6 12 93.0 110.8 75 71.2 52 154.6 14

12 3,5-diCQA 514.8 352.8 190.6 11 190.6 178.6 44 134.8 9

13 3,4-diCQA 514.8 352.8 172.8 25 334.8 19 172.6 178.6 83 190.6 65 134.8 9

14 4,5-diCQA 514.8 352.8 202.6 17 172.6 17 172.6 178.6 70 190.6 31 134.8 6

15 CFQA 528.8 366.8 352.8 22 192.8 13 192.6 172.6 13 133.6 11

16 CFQA 528.8 366.8 172.6 44 352.8 38 172.6 192.6 65

17 CFQA 528.8 352.8 266.8 95 172.6 27 172.6 178.6 61 190.6 30 134.6 8

132

Table 3. Colombia Supremo Roasted Bean Water Extraction Interpretations MS 1 MS 2 MS 3

No Compd. base peak Secondary peak base peak secondary peak Parent ion m/z m/z int m/z int m/z m/z int m/z int m/z int

1 Sucrose 377.0 340.8 - 178.6 160.8 28 142.8 23 112.8 20

2 QA 190.8 126.8 172.6 52 110.8 42 108.6 85.0 94

3 Pentose-QA 323.0 190.8 172.8 26 93.0 6 126.8 85.0 65 172.8 47 111.0 27

4 QA de. 364.8 190.6 172.6 5 -

5 QA ep. 190.8 110.8 172.6 24 67.4 -

6 QA de. 364.8 172.6 110.8 97 323 50 110.8 -

7 3-CQA ep. 353.0 178.6 134.8 33 172.8 16 134.8 -

8 3-CQA 352.8 190.6 178.8 46 134.8 13 126.8 172.8 61 110.8 48 85.0 37

9 5-CQA ep. 352.8 190.6 - 126.8 85.2 53 172.6 45 109.0 29

10 3-CQA ep. 353.0 190.6 178.6 75 134.8 27 172.6 93.0 54 126.8 44 110.8 25

11 5-CQA 352.8 190.6 - 85.0 172.6 87 126.8 83 93.0 78

12 3-FQA ep. 367.0 192.6 172.8 39 133.8 18 133.8 148.8 26 177.6 8

13 CQAL 335.0 178.6 160.8 51 134.8 37 134.8 -

14 3-FQA 367.0 192.8 133.8 19 133.8 148.8 21

15 4-CQA 352.8 172.8 178.8 63 190.8 19 93.0 111.0 78 154.6 35 71.4 31

16 5-FQA ep. 367.0 190.6 172.8 6 108.8 126.8 91 172.6 70 85.0 69

17 5-CoQA 337.0 190.6 172.8 5 172.6 110.8 88 126.8 47 85.2 33

18 3-CQAL 335.0 160.6 134.8 87 178.6 10 132.8 -

19 CQAL 335.0 160.6 178.8 100 210.8 78 132.8 -

20 5-FQA 367.0 190.6 - 85.2 172.8 66 126.8 58 93.0 42

21 CQAL 335.0 172.6 178.6 13 134.8 8 110.8 136.8 46 154.8 45 93.0 40

22 4-CQAL 335.0 160.6 134.8 26 290.8 17 132.8 -

23 CQAL ep. 335.0 160.6 172.8 51 132.8 11 93.0 110.8 49 71.2 47 154.6 13

24 4-FQA 367.0 172.6 192.6 12 93.0 110.8 54 154.8 27 81.2 16

25 diCQA ep. 515.0 340.8 280.8 86 190.8 77 280.8 296.8 53 254.8 19 322.8 7

26 diCQA ep. 515.0 340.8 322.8 62 280.8 59 296.8 280.8 60 254.8 10 324.0 9

27 3,5-diCQA 515.0 353.0 340.8 16 190.8 9 190.8 178.8 41 134.8 15

28 3,4-diCQA 515.0 353.0 172.8 26 178.8 21 172.8 178.6 77 190.6 53 134.8 31

29 4,5-diCQA 515.0 352.8 172.8 20 202.8 11 172.8 178.6 58 190.8 26 134.8 9

30 CFQA 529.0 367.0 172.6 39 349.0 26 172.6 192.6 25 133.8 9

31 CFQA 529.0 352.8 366.8 85 172.6 20 172.6 190.6 34 178.8 30 135.0 12

133

Table 4. Colombia Supremo Green Bean Water Extraction Interpretations MS 1 MS 2 MS 3

No Compd. base peak Secondary peak base peak secondary peak Parent ion m/z m/z int m/z int m/z m/z int m/z int m/z int

1 Sucrose 341.0 178.6 160.8 38 142.8 30 142.8 89.0 97 118.8 63 160.8 58

2 diCA 322.8 304.8 142.8 24 160.6 15 124.8 214.6 25 142.6 12 101.0 12

3 QA 190.6 126.8 172.8 49 110.8 32 108.8 99.0 28 71.2 7

4 Pentose-QA 322.8 190.6 172.6 66 110.8 8 172.4 93.0 96 110.8 61 126.8 45

5 QA de. 365.0 190.6 172.8 3 172.6 87.0 68 126.8 41 108.4 21

6 QA ep. 190.6 110.8 172.6 28 -

7 3-CQA ep. 352.8 190.8 178.6 86 134.8 24 172.8 93.0 81 136.8 24 126.8 13

8 3-CQA ep. 353.0 178.6 134.8 27 190.8 19 134.8 -

9 3-CQA 353.0 190.6 178.8 38 134.8 15 126.8 85.0 76 172.6 61 93.0 55

10 3-FQA ep. 367.0 334.8 160.6 52 178.6 49 172.8 178.8 94 160.6 82 134.8 74

11 5-CQA ep. 353.0 190.6 - 126.8 172.8 55 85.0 31 144.8 24

12 3-CQA ep. 352.8 190.6 178.6 88 134.8 23 93.0 172.6 64 111.0 38 126.8 19

13 5-CQA 352.8 190.6 178.8 4 126.8 172.6 66 93.0 64 85.0 44

14 M-3-CQ 367.0 160.6 192.6 11 132.8 11 132.8 105.0 35 116.6 4

15 3-CQA ep. 353.0 190.6 178.8 83 134.8 20 127.0 172.8 89 162.8 72 93.6 48

16 3-FQA ep. 367.0 192.6 172.6 42 133.8 15 133.8 148.8 27 177.6 16 116.8 5

17 CQAL 334.8 178.6 160.6 60 134.8 41 134.8 -

18 3-FQA 367.0 192.6 133.8 11 133.8 148.8 19

19 3-FQA ep. 367.0 334.8 192.8 31 160.6 26.6 178.6 160.6 44 134.8 31 254.8 26

20 4-CQA 353.0 172.6 178.6 58 134.8 13 93.0 110.8 84 154.8 30 71.4 29

21 5-FQA ep. 367.0 190.6 172.8 6 172.6 126.8 100 108.8 46 85.0 28

22 5-CoQA 337.0 190.6 162.8 9 178.8 6 85.0 126.8 63 172.8 44 108.8 44

23 3-CQAL 335.0 160.6 134.8 81 178.8 10 132.8 105.0 5

24 M-4-CQ 367.0 160.6 134.8 40 192.8 10 132.8 -

25 CQAL ep. 335.0 160.8 178.8 78 254.8 52 158.6 132.8 53

26 CQAL ep. 335.0 172.6 178.8 16 160.8 9 93.0 110.8 77 59.4 25 136.8 24

27 5-FQA 366.8 190.6 192.8 6 126.8 172.8 90 85.0 83 93.0 74

28 4-CQAL 334.8 160.8 134.8 30 178.8 23.8 132.8 -

29 CQAL ep. 335.0 160.6 172.8 42 178.6 7 132.8 -

30 4-FQA 367.0 172.6 192.6 8 93.0 110.8 46 71.2 24 154.6 17

31 diCQA ep. 515.0 340.8 280.8 96 190.6 69 296.8 280.8 59 254.8 12 178.8 7

32 diCQA ep. 515.0 352.8 190.8 52 341.0 32 190.6 178.6 12 134.8 10 172.8 4

33 3,5-diCQA 515.0 352.8 190.8 6 340.8 6 190.6 178.8 51 134.8 13 172.6 4

34 3,4-diCQA 515.0 352.8 172.6 25 178.6 20 172.6 178.6 53 190.8 37 134.8 15

35 CFQA 529.0 367.0 160.8 10 420.0 4 160.6 134.8 30 192.6 14 335.0 6

36 4,5-diCQA 515.0 352.8 172.8 16 202.6 14 172.8 178.8 63 190.8 29 134.8 13

37 CFQA 529.0 367.0 353.0 42 192.8 13 192.6 172.6 6 133.8 5

38 CFQA 529.0 367.0 172.8 56 352.8 42 172.8 192.6 69 110.8 6 119.8 6

39 CFQA 529.0 352.8 367.0 83 172.6 24 172.6 178.6 56 190.6 42 308.8 7

134

3.2. Interpretations of roasted bean data

As already published, a lot of reactions are occurring during the roasting period 7. We can easily find peak forming or increasing after roasting by comparing Figure 1 -A and

1-C. Figure 2 shows three different roasting compounds which showed significant peak increments; epimerization (Fig. 1-C-9), lactonisation (Fig. 1-C-18 and 1-C-22), and acyl-migration (Fig. 1-C-24).

O O HO OH HO OH O a O O OH O OH OH OH

HO HO OH OH

c b

O O HO HO OH O O O HO O O OH OH

OH HO OH OH

Figure 2. Chemical transformations occur during roasting process: (a) epimerization, (b) lactonisation, (c) acyl-migration

3.3. Differences between water and aqueous methanol (70 %) extracts

As shown in Figure 1 , meaningful composition differences are found between water and aqueous methanol (70 %) extracts data; 5-FQA and three di-CQAs give significant differences, together with other compounds.

135

Table 5. Relative ratios against 5-CQA (100 %) of major compounds Roasted / Roasted / Compd Green / Water Green / MeOH Water MeOH 5-FQA 12.78 27.09 22.75 36.08 3,5-diCQA 5.47 10.18 2.10 4.42 3,4-diCQA 2.85 8.29 2.71 6.61 4,5-diCQA 4.83 17.05 2.49 8.86 M-3-CQ - - 0.77 10.84 3-CQAL - - 27.34 39.28 4-CQAL - - 24.23 26.92

Table 5 showed seven different compounds which show decided intensity differences.

The intensities were measured in the extracted ion chromatograms (EICs) of 353 (5-

CQA, standard, 100 %), 367 (5-FQA and M-3-CQ), 515 (di-CQAs), and 335 (CQALs).

In all seven shown compounds, relative intensities against 5-CQA show meaningful increase in methanol extracts than water extracts.

4. Conclusion

In conclusion, we analyzed Arabica Colombia Supremo coffee beans with liquid chromatography / tandem mass spectrometry in four different conditions; water extracted green bean, methanol extracted green bean, water extracted 240 ℃ 15 min roasted bean, and methanol extracted 240 ℃ 15 min roasted bean. We interpreted polyphenolic compounds in Total MS spectra, and reconfirmed some published chemical reactions during coffee roasting steps. Most importantly, we verified for the first time, the effect of coffee extracting solvents between water and methanol by comparing major compounds’ detected intensities of relative quantities against 5-CQA.

136

We expect this result would help further research directly reflect the real coffee consumption in order to fill up a gap between academic and practical researches.

Acknowledgements

Excellent technical support by Anja Müller and providing coffee beans from Lloyd

Caffee are acknowledged.

References

1. Kuhnert, N.; Jaiswal, R.; Eravuchira, P.; El-Abassy, R.M.; von der Kammer, B.; Materny, A. Scope and limitations of principal component analysis of high resolution LC-TOF-MS data: the analysis of the chlorogenic acid fraction in green coffee beans as a case study. Analytical Methods 2011, 3, 144-155.

2. Balyaya, K.J.; Clifford, M.N. Chlorogenic acids and caffeine contents of monosooned Indian Arabica and Robusta coffees compared with wet and dry processed coffees from the same geographic area. Colloq. Sci. Int. Cafe, [C. R. ] 1995, 16th, 316-325.

3. Moon, J.; Yoo, H.S.; Shibamoto, T. Role of Roasting Conditions in the Level of Chlorogenic Acid Content in Coffee Beans: Correlation with Coffee Acidity. J. Agric. Food Chem. 2009, 57, 5365-5369.

4. Moon, J.; Shibamoto, T. Role of Roasting Conditions in the Profile of Volatile Flavor Chemicals Formed from Coffee Beans. J. Agric. Food Chem. 2009, 57, 5823-5831.

5. Moon, J.; Shibamoto, T. Formation of Volatile Chemicals from Thermal Degradation of Less Volatile Coffee Components: Quinic Acid, Caffeic Acid, and Chlorogenic Acid. J. Agric. Food Chem. 2010, 58, 5465-5470.

6. Gross, G.; Jaccaud, E.; Huggett, A.C. Analysis of the content of the diterpenes cafestol and kahweol in coffee brews. Food and Chemical Toxicology 1997, 35, 547- 554.

7. Jaiswal, R.; Matei, M.F.; Golon, A.; Witt, M.; Kuhnert, N. Understanding the fate of chlorogenic acids in coffee roasting using mass spectrometry based targeted and non- targeted analytical strategies. Food & Function 2012, 3, 976-984.

8. Ky, C.L.; Louarn, J.; Dussert, S.; Guyot, B.; Hamon, S.; Noirot, M. Caffeine, trigonelline, chlorogenic acids and sucrose diversity in wild Coffea arabica L. and C- canephora P. accessions. Food Chem. 2001, 75, 223-230.

137

9. Farah, A.; De Paulis, T.; Trugo, L.C.; Martin, P.R. Effect of roasting on the formation of chlorogenic acid lactones in coffee. J. Agric. Food Chem. 2005, 53, 1505-1513.

10. Semmelroch, P.; Grosch, W. Analysis of roasted coffee powders and brews by gas chromatography-olfactometry of headspace samples. Lebensmittel-Wissenschaft and Technologie 1995, 28, 310-313.

12. Matei, M.F.; Jaiswal, R.; Kuhnert, N. Investigating the Chemical Changes of Chlorogenic Acids during Coffee Brewing: Conjugate Addition of Water to the Olefinic Moiety of Chlorogenic Acids and Their Quinides. J. Agric. Food Chem. 2012, 60, 12105-12115.

13. Kemsley, E.K.; Ruault, S.; Wilson, R.H. Discrimination between Coffea-Arabica and Coffea-Canephora Variant Robusta Beans using Infrared-Spectroscopy. Food Chem. 1995, 54, 321-326.

14. Briandet, R.; Kemsley, E.K.; Wilson, R.H. Discrimination of Arabica and Robusta in instant coffee by Fourier transform infrared spectroscopy and chemometrics. J. Agric. Food Chem. 1996, 44, 170-174.

15. Fujioka, K.; Shibamoto, T. Chlorogenic acid and caffeine contents in various commercial brewed coffees. Food Chem. 2008, 106, 217-221.

16. Iwai, K.; Kishimoto, N.; Kakino, Y.; Mochida, K.; Fujita, T. In vitro antioxidative effects and tyrosinase inhibitory activities of seven hydroxycinnamoyl derivatives in green coffee beans. J. Agric. Food Chem. 2004, 52, 4893-4898.

17. Clifford, M.N.; Johnston, K.L.; Knight, S.; Kuhnert, N. Hierarchical scheme for LC- MSn identification of chlorogenic acids. J. Agric. Food Chem. 2003, 51, 2900-2911.

18. Kuhnert, N.; Jaiswal, R.; Matei, M.F.; Sovdat, T.; Deshpande, S. How to distinguish between feruloyl quinic acids and isoferuloyl quinic acids by liquid chromatography/tandem mass spectrometry. Rapid Communications in Mass Spectrometry 2010, 24, 1575-1582.

19. Jaiswal, R.; Kuhnert, N. Hierarchical scheme for liquid chromatography/multi-stage spectrometric identification of 3,4,5-triacyl chlorogenic acids in green Robusta coffee beans. Rapid Communications in Mass Spectrometry 2010, 24, 2283-2294.

138

3.3.2. Research on Brewing Step with Different Methods

Brewing Less- and Enhanced-caffeinated Coffee by Using Cold Brew

Method

Seung-Hun Lee, Nikolai Kuhnert*

Department of Life Sciences and Chemistry, Jacobs University Bremen, 28759 Bremen,

Germany

*Correspondence to:

Prof. Dr. Nikolai Kuhnert,

Department of Life Sciences and Chemistry

Jacobs University Bremen gGmbH

Campus Ring 1

28759 Bremen, Germany

Tel: +49 421 200 3120

Fax: +49 421 200 3102

E-mail: [email protected]

139

Abstract

Background: Caffeine is the most attractive compound in coffee on account of its remarkable pros and cons for human health. In contemporary human diet, coffee is the main source for caffeine intake, so that controlling daily caffeine intake became an issue.

The only caffeine control method existing is decaffeination and it needs a lot of starting setup expense so that it is allowed only for a major companies.

Methods: A modern coffee brewing method, cold brew method, was used and the coffee was analyzed in chronological order to interpret sequential changes of extracted compounds. Roasted coffee powder 50 g was used for extracting cold brew coffee and samples were collected every 50 mL until 24 samples. HPLC-MS were used for interpreting compounds and UV integration was used for measuring the quantitative data.

Results: Total 26 compounds composed of chlorogenic acids of mono- and di-acylated p-coumaric acid derivatives and caffeine in cold brewed coffee were interpreted. The most important content change was caffeine in normalized coffee concentration.

Continuous increasing of caffeine content was detected while commercially used cold brewing method; until 10 times of water to the coffee powder used.

140

Conclusions: Without any additional expense, max. ± 25 % caffeine content control method is developed with cold-brew method. The LeeN-Caffeination method would be expected as a standard content ratio control method in many different ways; different compound control other than caffeine, different brewing method research other than dripping cold-brew or different extracting materials other than coffee.

Keywords: Coffee; Cold brew coffee; Caffeine control; LeeN Caffeination; LC-MS

141

1. Introduction

It has been continuously emphasized the importance of coffee in 21 st century. Although the old myth started to spread from the book [Uncommon Grounds: The History of

Coffee and How it Transformed our World] by Mark Pendergrast, “Coffee is the second most valuable exported legal commodity on earth after oil.”, 1 is now known as false, 2 at least rapid growth of coffee market became an important economic issue.

In spite of short history, both words, polyphenol and chlorogenic acids became well- known to public by advertisements of global coffee companies to make coffee as a healthy beverage. 3 In 2003, HPLC-tandem MS allowed us to distinguish the regioisomers of caffeoylquinic acids. 4 After the research, interpretation of chlorogenic acids became an important part of analyzing the differences in all the coffee processes such as different origins, species, roasting degrees and brewing methods. On the other hand, although there is still an animated controversy, caffeine is known to public as a restriction-needed compound for consistent coffee consumption. 5 Most restrictions of coffee per day have grounds on the calculation of caffeine intake per day. So the decaffeinated coffee market is growing rapidly.

The movement of caffeine control in coffee started from 1903 by Ludwig Roselius in

Bremen, Germany. 6 Since then, a lot of decaffeination methods have been used for the people who do not want to consume large amount of caffeine. Starting of benzene extracting method by Ludwig Roselius, dichloromethane extraction 7 is used on account of its low boiling point, and ethyl acetate extraction 8 is used with the name of natural decaffeination because it exists in natural aromatic fruits. Supercritical carbon dioxide 9

142

is also used as an extracting solvent, not widely used owing to its high-cost, and the most widely used method is the so called Swiss Water Process 10 for using water as an extracting solvent. However, all these decaffeination methods have fundamental limitation of high facility cost for the customers to control caffeine in their own preference.

Actually, there’re not enough selections for customers on coffee from farm to table.

Coffee production is restricted in so called “Coffee Belt” by the weather and soil. Wet, semi-dry and dry coffee processes to make green coffee bean from coffee cherry could only be done near the farm by distance and size of facilities. Uniform roasting needs at least more than 1 kg roaster which costs more than 10,000 Euro. The only choice allowed for coffee consumers is to select the brewing method. Espresso, electric coffee maker and instant coffee are the major part of brewing methods. However, as coffee gourmets appear, more sincere brewing processes have been growing started from

Europe and USA, and spread from Korea and Japan; a variety of hand-brewing methods and cold-brewing methods. 11,12

Cold brewing method has two different ways to brew coffee; steeping cold brew and dripping cold brew. As the name cold brewing method says, people use 0 oC ~ RT water to brew coffee. In the steeping cold brew method, water is poured into ground coffee, the water/ground coffee mixture is stored for more than 12 hours and it is filtered to make concentrated coffee. However, dripping cold brew method, as known as cold drip or dutch brewing coffee, is a continuous method. It needs three parts, water reservoir, ground coffee column and coffee bottle. From coffee reservoir water drops with the

143

speed of 1 drop per 1 - 5 seconds to pass through ground coffee column filled with ground coffee to extract coffee and arrives at the bottom coffee bottle only with the power of gravity. The commercial ratio of ground coffee versus used water is 1:5 to

1:10. The structure of dripping cold brew is shown in Figure 1 .

Figure 1 . Structure of dripping cold brew

In this research, dripping cold brew method is used for brewing coffee samples, and the samples are divided in sequential order. Cold brewed kinetic samples were analyzed using HPLC/tandem MS to interpret chlorogenic acids and caffeine content changes.

Quantification of caffeine content in kinetic samples was processed in order to separate

Less-caffeinated and Enhanced-caffeinated coffee samples to achieve Less- and

144

Enhanced-caffeine control method by Seung-Hun Lee and Nikolai Kuhnert (LeeN- caffeination).

2. Experimental

The experimental methods of HPLC and LC/MS n are followed by previous publications of our group research. 13-15

2.1. Chemicals, Coffee Beans and Brewing Materials

All the chemicals (Analytical grade) were purchased from Sigma-Aldrich (Germany).

Roasted Brazil Arabica coffee bean was purchased from Lloyd Caffee in Bremen,

Germany. Product name ‘Juliet’s Tears’, 500 mL cold brewing set, was provided by

Caferomeo (Seoul, Korea). Water for brewing coffee, product name ‘Mineral Wasser’, was purchased from EDEKA in Bremen.

2.2. Brewing of Coffee Beans

50 g of roasted Brazil Arabica coffee bean was ground to fine powder and put in the coffee column on the Kalita 56 mm coffee filter. After flatting with hands patting, another coffee filter is put on the coffee powder for the even distribution of dropping water. After the coffee column is filled, water drops with the speed of one drop per three seconds from water reservoir to the coffee column. The water extracts coffee and passes through the coffee column to the bottom sampling bottle. The sampling bottle is changed in every 50 mL. The sampling is continued until 24 th sample, 1200 mL in total extraction has been collected. The 24 consecutive samples are shown in Figure 2 .

145

Figure 2 . Cold brew coffee samples 1 ~ 24

The extracts were stored at 4 ℃ until required, aged at room temperature, filtered through a membrane filter, diluted to 5 % with 70 % aqueous methanol and used for

LC/MS. The concentrations of the extracted liquid were measured by evaporating solvent at 100 ℃ for 30 min using aluminum boat to measure boat, boat with solution before evaporation and boat with solid after evaporation.

2.3. High Resolution MS

The LC equipment (Agilent 1100 series, Karlsruhe, Germany) comprised a binary pump, an autosampler with a 100 µL loop, and a diode-array detector (DAD) with a light-pipe flow cell (recording at 320 and 254 nm and scanning from 200 to 600 nm). This was interfaced with a MicroTOF Focus mass spectrometer (Bruker Daltonics, Bremen,

146

2.4. LC/MS n

The LC equipment comprised a binary pump, an autosampler with a 100 µL loop, and a diode-array detector (DAD) with a light-pipe flow cell (recording at 320 and 254 nm and scanning from 200 to 600 nm). This was interfaced with an ion-trap mass spectrometer fitted with an ESI source (Bruker Daltonics HCT ultra, Bremen, Germany) operating in full scan, auto MS n mode (smart fragmentation) using a ramping of the collision energy. Maximum fragmentation amplitude was set to 1 V, starting at 20 % and ending at 200 %. MS operating conditions (negative mode) had been optimized using 5-caffeoylquinic acid with a capillary temperature of 365 ℃, a drying gas flow rate of 10 L / min, and a nebulizer pressure of 10 psi.

2.5. HPLC

Separation was achieved on a 150 × 3 mm i.d. column containing diphenyl 5 µm, with a

5 mm × 3 mm i.d. guard column (Varian, Darmstadt, Germany). Solvent A was water / formic acid (1000:0.005 v/v) and solvent B was methanol. Solvents were delivered at a total flow rate of 500 µL / min. The gradient profile was from 10 to 80 % B linearly in

70 min followed by 10 min a return to 10 % B and 10 min isocratic to re-equilibrate.

3. Results and Discussion

3.1. Interpretations of chlorogenic acids

3.1.1. Molecular formula analysis of high resolution MS data

MicroTOF was used to obtain exact molecular weight to determine chemical formula of chlorogenic acids in cold brew coffee. Figure 3 is the base peak chromatogram of the first 50 mL samples (sample 1) of cold brew coffee by LC-TOFMS.

147

Figure 3. Base peak chromatogram of cold brew coffee by LC-TOFMS

The numbering of the compounds is shown in supporting information ( supplementary table 1 , supplementary figure 1 and supplementary figure 2 ). Following Table 1 shows average molecular weight of different chlorogenic acids and caffeine in positive mode found in the three repeated cold brew sample 1.

148

Table 1. High resolution mass data in the first 50 mL samples of cold brew coffee

No. RT Name Mol. Formula Exp. m/z (M+H) Theor. m/z (M+H) Error (ppm)

4ui 20.0 Caffeoylquinic acid (ep) C16 H18 O9 355.1017 355.1024 1.9

4ui 21.5 Caffeoylquinic acid (ep) C16 H18 O9 355.1009 355.1024 4.1

4c 23.9 3-Caffeoylquinic acid C16 H18 O9 355.1016 355.1024 2.1

4ui 26.6 Caffeoylquinic acid (ep) C16 H18 O9 355.1014 355.1024 2.7

4ui 27.5 Caffeoylquinic acid (ep) C16 H18 O9 355.1017 355.1024 1.9

4e 29.3 5-Caffeoylquinic acid C16 H18 O9 355.1021 355.1024 0.7

Caf 30.3 Caffeine C8H10 N4O2 195.0883 195.0877 -3.3

3c 30.3 3-p-Coumaroylquinic acid C16 H18 O8 339.1050 337.0918 7.2

1d 30.6 4-Caffeoylshikimic acid C16 H16 O8 337.0899 369.1180 5.6

6c 31.4 3-Caffeoylquinic acid methyl ester C17 H20 O9 369.1173 369.1180 1.9

5c 33.4 3-Feruloylquinic acid C17 H20 O9 369.1169 355.1024 3

4d 34.8 4-Caffeoylquinic acid C16 H18 O9 355.1015 339.1074 2.4

2c 35.9 3-Caffeoyl-1,5-quinide C16 H16 O8 337.0896 337.0918 6.5

6d 36.3 4-Caffeoylquinic acid methyl ester C17 H20 O9 369.1163 339.1074 4.6

3e 36.7 5-p-Coumaroylquinic acid C16 H18 O8 339.1059 339.1074 4.6

1e 37.4 5-Caffeoylshikimic acid C16 H16 O8 337.0903 369.1180 4.4

5e 38.6 5-Feruloylquinic acid C17 H20 O9 369.1165 337.0918 4.1

2d 39.4 4-Caffeoyl-1,5-quinide C16 H16 O8 337.0903 369.1180 4.4

3d 41.1 4-p-Coumaroylquinic acid C16 H18 O8 339.1064 337.0918 3.1

2a 41.3 1-Caffeoyl-1,5-quinide C16 H16 O8 337.0902 337.0918 4.7

9u 43.6 Feruloyltrimethoxycinnamoylquinic acid C29 H33 O13 589.1918 589.1916 -0.4

5d 43.6 4-Feruloylquinic acid C17 H20 O9 369.1165 369.1180 4.1

7ce 47.4 3,5-Dicaffeoylquinic acid C25 H24 O12 517.1305 517.1341 6.9

7cd 49.7 3,4-Dicaffeoylquinic acid C25 H24 O12 517.1323 517.1341 3.4

7de 51.3 4,5-Dicaffeoylquinic acid C25 H24 O12 517.1331 517.1341 1.8

10u 55.9 Feruloylsinapoylquinic acid C28 H31 O13 575.1767 575.1759 -1.4

3.1.2. Interpretation methods of chlorogenic acids by their fragmentation pattern

Interpretation methods of chlorogenic acids by LC/MS n have been already studied especially in hydroxycinnamic acid derivatives; caffeoyl quinic acids, 4 feruloylquinic acids, 16 and others. 17

149

Figure 4 . Total MS spectra of cold brew coffee samples

Compounds were interpreted in four different spectra as shown in Supplementary table 2 ~ 5 with their Total MS data in Figure 4. We can find that the number of compounds and integrations decrease as sample changes. We will discuss it in 3.2.

Results on brewing kinetics .

3.2. Results on brewing kinetics

3.2.1. Brewing kinetics on concentration

150

It is easy to predict that the concentration would decrease while brewing is progressing because the compounds are extracting out, and the real concentration showed as expected. The average solid content in triplicated extraction samples with standard deviation are shown in Figure 5.

Figure 5. Brewing kinetics on the total solid content

Common cold-brewing method utilizes 50 g of coffee and 10 times of water (500 mL) to get about 400 mL of concentrated coffee (around 50 to 100 mL of water are conserved in the coffee column with coffee). Therefore, the first 8 samples are commercially used, and others are discarded and even if they are used, the actual

151

content would not influence the total compound ratio because of its low concentration.

In our experimental data, more than 85 % of solid content was until 8 th sample in total

24 samples.

3.2.2. Brewing kinetics on caffeine and 5-CQA

Positive mode extracted ion chromatogram (EIC) at m/z 195 ± 0.5 Da and negative mode EIC at m/z 353 ± 0.5 Da were used for interpreting caffeine and 5-CQA, respectively and UV spectra at 280 nm was used for quantifying both caffeine and 5-

CQA. As following the concentration data, it was observed that both compounds showed decreasing intensity by kinetics. The average caffeine and 5-CQA integrations in triplicated extractions with standard deviations are shown in Figure 6.

152

Figure 6. Brewing kinetics on caffeine and 5-CQA

3.2.3. Brewing kinetics on caffeine and CQAs in normalized concentration

By using kinetic peak integrations, conclusive data could not be obtained as they only showed continuous decreasing graph influenced by kinetic concentration changes as shown in Figure 5. In Figure 7, concentration normalized caffeine graph show increasing contents in the same concentration until sample 7, which is almost the limit of commercially used extracting volume.

153

Figure 7. Brewing kinetics on normalized caffeine content

The main achievement of this research is to obtain the ability to control caffeine content by brewing. For achieving this, the quantification data of caffeine is required.

Quantification has been done comparing with reference caffeine purchased from Sigma-

Aldrich with integration data of UV spectra. If we regard commercially consumed coffee as 1 % solution, the caffeine content in Figure 7 gives 62.1 mg (Sample 1) to

97.2 mg (Sample 7) per one cup of coffee we drink. The average caffeine content in 1% concentration of 100 mL coffee samples is 78.7 mg (commercial average, sample 1 ~ 8) which is 26.7 % higher than the same concentration coffee of Sample 1 and 23.5 % lower than that of Sample 7.

154

There has been a controversy in this brewing method that cold brewing extracts less caffeine because of its temperature versus cold brewing extracts more caffeine because of its brewing time. Both claims have validity with real data and are still in controversy.

We can find the reason in Figure 6, this method has high standard deviations so that the exact amount cannot be guaranteed in the range of two to three folds. It can be speculated that the high error in this method may be due to the uneven water distribution for the ground coffee. On the top of coffee column, the water drops only in the center of the column, so fair distribution of water cannot be guaranteed especially for the early stages. Furthermore, there is also the possibility for the formation of water channels that may affect the even distribution of water. In this research, we tried to get certain data by triplicating and averaging all the processes.

To sum up, in Figure 7, we can get less- and enhanced caffeinated undiluted coffee with maximum caffeine content difference of 56.5 % from sample 1 to sample 7.

4. Conclusion

In conclusion, 24 kinetic samples of Brazil roasted coffee extracts by using cold brewing method were analyzed in this study. LC-TOFMS and LC-IontrapMS were used for analyzing the extracts to interpret chlorogenic acids and caffeine in each sample.

Most importantly, it was found that caffeine content in 100 mL coffee could be controlled from 62.1 mg (Sample 1) to 97.2 mg (Sample 7) when we drink coffee in 1 % concentration.

155

This is the first caffeine controlling method after the decaffeination methods were discovered. And it is the only method that caffeine can be controlled in the final stage of coffee process. All the decaffeination methods have had to be done in industrial scale because of its huge cost, but this LeeN-caffeination could be done by not only in café but also in ordinary houses. Although LeeN-caffeination has the limitation that Less- caffeination and Enhanced-caffeination are only around 25 % of normal caffeine content, this difference is enough for some groups. For example, the recommended caffeine intake per day is 400 mg for normal adult and 300 mg for pregnant women in most of the countries. The 25 % caffeine reduced coffee allows the pregnant group to drink coffee the same as what they consumed before the pregnancy which might reduce some part of their pregnancy stress. And the residues can be used for the people who need to work actively for the daytime.

A lot of further research in this matter can be expected starting from this research; kinetic research of espresso and filter coffee machine by changing samples in seconds, higher- or lower-temperature brewing with cold brewing method and another cold brewing method; steeping cold brew studies.

5. Acknowledgements

Excellent technical support by Anja Müller, Abhinandan Shrestha for his proofreading and providing cold brewing machine from Café Romeo (Seoul, Korea) are acknowledged.

156

6. Author Disclosure Statement

No competing financial interests exist.

References

1. Pendergrast M. Uncommon grounds: The history of coffee and how it transformed our world. Basic Books; 1999.

2. FAO. Food and agriculture organization of the united nations. www.fao.org .

3. Quideau S, Deffieux D, Douat-Casassus C, Pouysegu L. Plant polyphenols: Chemical properties, biological activities, and synthesis. Angewandte Chemie-International

Edition . 2011;50(3):586-621.

4. Clifford MN, Johnston KL, Knight S, Kuhnert N. Hierarchical scheme for LC-MSn identification of chlorogenic acids. J Agric Food Chem . 2003;51(10):2900-2911.

5. Nawrot P, Jordan S, Eastwood J, Rotstein J, Hugenholtz A, Feeley M. Effects of caffeine on human health. Food Additives and Contaminants Part A-Chemistry Analysis

Control Exposure & Risk Assessment . 2003;20(1):1-30.

6. Meyer JF, Jr., Roselius L, Wimmer K, inventors; Germany, assignee. Removing caffeine from coffee. 1908.

7. Zacek Z, inventorDecaffeinated green coffee. 1960.

157

8. Morrison LR, Jr., Elder MN, Phillips JH, inventors; Procter and Gamble Co., assignee. Recovery of noncaffeine solubles in an extract decaffeination process. 1982.

9. Zosel K, inventor; Studiengesellschaft Kohle m.b.H., assignee. Decaffeinizing coffee.

1979.

10. Fischer A, Kummer P, inventors; Coffex A.-G., assignee. Decaffeinization of raw coffee. 1980.

11. Bentz HW, inventor; Melitta-Werke A.-G., assignee. Filters. 1938.

12. Tarasch LC, inventorExtracting and preserving macerated coffee. 1957.

15. Kuhnert N, Jaiswal R, Eravuchira P, El-Abassy RM, von der Kammer B, Materny

A. Scope and limitations of principal component analysis of high resolution LC-TOF-

MS data: The analysis of the chlorogenic acid fraction in green coffee beans as a case study. Analytical Methods . 2011;3(1):144-155.

16. Kuhnert N, Jaiswal R, Matei MF, Sovdat T, Deshpande S. How to distinguish between feruloyl quinic acids and isoferuloyl quinic acids by liquid

158

chromatography/tandem mass spectrometry. Rapid Communications in Mass

Spectrometry . 2010;24(11):1575-1582.

17. Jaiswal R, Kuhnert N. Hierarchical scheme for liquid chromatography/multi-stage spectrometric identification of 3,4,5-triacyl chlorogenic acids in green robusta coffee beans. Rapid Communications in Mass Spectrometry . 2010;24(15):2283-2294.

159

Supplementary table 1. List of chlorogenic acids

No Name Abb Q Rx R’ 1 Caffeoylshikimic acid CSA 1db2 C H 2 Caffeoyl-1,5-quinide CQL 1la5 C la 3 p-Coumaroylquinic acid CoQA - Co H 4 Caffeoylquinic acid CQA - C H 5 Feruloylquinic acid FQA - F H

6 Caffeoylquinic acid methyl ester M-CQ - C CH 3 7 diCaffeoylquinic acid diCQA - diC H 8 Caffeoylferuloylquinic acid CFQA - C, F H 9 Feruloyltrimethoxycinnamoylquinic acid FTMQA - TM, F H 10 Feruloylsinapoylquinic acid FSQA - S, F H

* ‘a’, ‘c’, ‘d’ and ‘e’ stand for stereochemical positions of quinic acid; 1, 3, 4 and 5, respectively. ‘u’ is used for unconfirmed isomeric positions and ‘i’ is used for stereoisomers. The stereoisomeric position numbering is shown in Supplementary figure 1 . Some examples of each numberings are shown in Supplementary figure 2 .

160

O RaO OR'

1 2 6 3 5 RcO 4 ORe ORd

Supplementary figure1. Stereoisomeric positions of quinic acid

161

O OH O OH HO OH HO O O O OH HO O O OH O OH O OH O OH OH HO HO 1c 2a 3c

O HO OH OH OH O HO O O O O O O OH HO HO HO O O OH OH OH OH OH OH HO 4e 5d 6c

O O HO OH HO OH O O O O HO OH O OH O O O O OH OH HO OH HO OH

7ce 8ec O O HO OH HO OH O O O O HO O O OH O O O OH HO O

O O 9ed 10cd O O O OH

Supplementary figure2. Chemical structures of chlorogenic acids with numbering

162

Supplementary table 2. Chlorogenic acids in Brazil cold brew coffee sample 1 MS 1 MS 2 MS 3

No Compd. Parent base peak Secondary peak base peak secondary peak ion m/z m/z int m/z int m/z m/z int m/z int m/z int

4ci 3-CQA de 352.95 190.8 178.9 85 135 25 93.2 170.8 39 111.1 31 172.8 31

4ci 3-CQA de 352.92 178.8 135 30 190.8 21 134.9 135.9 4 133.9 2 137 1

4c 3-CQA 352.93 190.8 178.9 45 134.9 11 127 85.3 75 93.2 71 172.9 58

4ei 5-CQA de 352.93 190.9 191.8 8 265.9 2 126.9 172.9 81 109.1 51 87.3 39

4ci 3-CQA de 352.95 190.9 178.8 73 134.9 17 93.2 111 57 172.9 51 170.9 27

4e 5-CQA 352.96 190.8 191.8 6 178.9 5 127 172.8 79 85.4 75 93.2 47

5ci 3-FQA de 367.01 192.8 172.9 28 331 13 133.9 148.9 28 117.1 4 134.9 3

5ci 3-CQA de 352.95 190.8 265.9 53 178.9 38 85.3 126.9 79 172.9 49 170.9 39

1c 3-CSA 334.89 178.8 160.9 63 134.9 29 134.9 135.9 5 134 2 107.1 1

4ui CQA de 352.94 190.8 265.9 62 172.9 26 127 93.2 98 85.2 71 172.9 22

5c 3-FQA 366.97 192.8 134 11 193.8 10 133.9 148.9 22 134.9 4 117 4

3ei 5-pCoQA de 336.92 190.9 191.8 8 334.9 4 126.9 172.8 96 85.3 42 87.3 38

4d 4-CQA 352.93 172.9 178.9 57 190.8 12 93.2 111.1 52 154.9 24 109.1 15

4di 4-CQA de 352.94 172.9 190.9 68 178.9 58 93.2 111.1 55 154.9 48 71.5 32

5ei 5-FQA de 366.97 190.8 191.8 5 172.9 4 127 172.9 66 144.9 31 85.3 23

3e 5-pCoQA 336.91 190.8 191.8 9 334.9 7 127 172.9 68 85.3 47 93.2 47

2c 3-CQAL 334.91 160.8 134.9 75 190.9 13 132.9 117 3 134 1 - -

4di 4-CQA de 352.97 190.9 172.9 97 533.1 37 93.2 111.1 23 - - - -

1e 5-CSA 334.92 172.8 178.9 25 160.9 19 111 93.3 45 154.9 38 71.6 37

5e 5-FQA 366.96 190.8 191.8 8 192.8 5 127 172.9 97 85.3 84 93.2 61

2d 4-CQAL 334.92 160.9 178.8 23 135 22 132.9 119 10 158.8 8 134 2

2a 1-CQAL 334.94 160.9 172.9 56 178.8 34 132.9 ------

4d 4-FQA 366.98 172.9 192.8 9 173.8 7 93.2 111 72 71.5 28 154.9 16

4di 4-FQA de 366.98 172.9 190.8 38 347.9 37 111 93.3 75 97.2 58 99.1 22

1ei 5-CSA de 334.91 172.8 314.8 20 332.9 13 93.2 59.5 95 111.2 94 128.1 84

7ce 3,5-diCQA 515.11 352.9 353.9 21 190.9 9 190.8 178.8 52 135 9 191.8 8

7cd 3,4-diCQA 515.11 352.9 172.9 27 178.9 16 172.9 178.8 64 190.8 58 135 16

7de 4,5-diCQA 515.09 352.9 340.9 28 172.9 19 280.9 296.9 93 297.9 26 254.9 13

8u C-4-FQA 529.12 367 172.9 46 352.9 37 172.8 192.8 42 190.9 10 133.9 6

163

Supplementary table 3. Chlorogenic acids in Brazil cold brew coffee sample 5 MS 1 MS 2 MS 3

No Compd. Parent base peak Secondary peak base peak secondary peak ion m/z m/z int m/z int m/z m/z int m/z int m/z int

4ci 3-CQA de 352.94 190.9 178.9 80 134.9 19 93.2 170.9 33 172.9 28 111.1 23

4ci 3-CQA de 352.93 178.9 135 23 172.9 12 134.9 133 11 135.9 10 134 5

4c 3-CQA 352.97 190.8 178.8 45 191.8 10 127 172.8 88 85.3 81 111 68

4ei 5-CQA de 352.93 190.8 191.8 6 172.9 3 127 172.9 83 144.8 35 85.3 31

4ei 5-CQA de 352.93 190.8 134.9 6 265.9 9 127 93.1 91 85.3 86 111 78

4e 5-CQA 352.94 190.9 191.8 7 214.9 3 172.8 127 88 85.3 85 93.2 69

6c M-3-CQ 366.98 160.9 192.8 16 134.9 12 132.9 ------

4ci 3-CQA de 352.93 190.8 178.8 21 265.9 7 127 172.8 72 111 62 93.2 58

1c 3-CSA 334.9 178.8 160.9 60 134.9 31 135 135.8 15 134 2 - -

4di 4-CQA de 352.92 190.9 265.9 29 172.9 27 172.8 127 84 108.9 62 170.9 37

5c 3-FQA 366.97 192.8 193.8 13 133.9 10 133.9 148.9 27 117 4 190.8 3

4d 4-CQA 352.93 172.9 178.8 48 190.8 13 93.2 111.1 62 154.9 21 59.7 17

4di 4-CQA de 352.93 172.9 190.8 56 178.9 44 93.3 170.8 24 115.1 23 111.1 18

5ei 5-FQA de 366.93 190.8 191.7 5 172.9 5 126.9 172.9 95 144.9 33 87.2 30

2c 3-CQAL 334.9 160.9 135 87 190.8 26 132.9 93.3 37 - - - -

6d M-4-CQ 367.02 160.9 134.9 24 161.8 13 132.9 ------

1e 5-CSA 334.92 172.8 178.8 29 160.9 23 111.1 93.2 61 154.8 38 71.5 36

5e 5-FQA 367 190.8 191.8 9 192.9 4 172.8 127 78 85.3 52 93.2 41

2d 4-CQAL 334.89 160.9 178.9 29 135 22 132.9 134 5 116.9 4 158.6 3

3d 4-pCoQA 336.96 172.8 334.9 24 173.8 12 111.1 93.2 93 154.9 71 71.5 49

2a 1-CQAL 334.93 160.9 172.9 46 317.8 24 132.9 ------

5d 4-FQA 367 172.9 192.9 12 173.8 7 93.2 111.1 59 71.5 33 154.9 17

7ce 3,5-diCQA 515.07 352.9 353.9 15 190.9 14 190.8 178.8 45 134.9 9 191.6 3

7cd 3,4-diCQA 515.1 353 172.9 25 334.9 18 178.9 172.8 83 190.8 54 135 21

7de 4,5-diCQA 515.09 352.9 202.8 17 172.9 15 172.9 178.9 69 190.8 27 134.9 13

164

Supplementary table 4. Chlorogenic acids in Brazil cold brew coffee sample 10 MS 1 MS 2 MS 3

No Compd. Parent base peak Secondary peak base peak secondary peak ion m/z m/z int m/z int m/z m/z int m/z int m/z int

4ci 3-CQA de 352.92 178.8 172.9 16 134.9 15 134.9 132.9 15 135.9 4 - -

4c 3-CQA 352.92 190.8 178.9 40 191.8 8 127 172.9 57 85.3 46 109.1 46

4ei 5-CQA de 352.91 190.8 191.8 7 350.9 4 126.9 172.9 94 85.3 46 109.1 27

4e 5-CQA 352.93 190.8 191.8 8 214.8 3 127 172.8 85 85.3 59 93.2 38

6c M-3-CQ 366.97 160.9 192.8 21 134.9 11 260 132.9 12 117.2 2 - -

4ci 3-CQA de 352.93 190.9 178.8 17 350.9 10 172.8 127 64 111.1 54 109.1 53

1c 3-CSA 334.98 178.9 160.9 52 135 30 134.9 108 4 135.9 4 117 3

5ci 3-FQA de 366.87 192.8 298.7 91 347.8 65 230.7 229.8 73 252.8 17 239.8 17

4di 4-CQA de 352.92 172.9 190.9 88 178.9 42 93.2 154.9 49 59.6 30 111.1 18

5c 3-FQA 366.98 192.8 334.9 17 134 12 133.9 148.9 22 190.8 5 117 3

4d 4-CQA 352.93 172.9 178.8 45 190.8 15 93.2 111.1 81 154.8 36 71.5 30

2ci 3-CQAL de 334.88 160.9 134.9 68 178.9 18 132.8 ------

2c 3-CQAL 334.91 160.9 134.9 70 178.9 11 132.9 117.1 2 - - - -

6d M-4-CQ 366.99 160.9 135 28 161.9 12 132.9 134 23 131.9 7 - -

1e 5-CSA 334.93 172.9 178.9 25 160.9 21 111 71.5 33 93.2 31 154.9 27

5e 5-FQA 366.97 190.8 191.8 8 192.8 5 172.9 85.3 96 93.2 77 127 74

2d 4-CQAL 334.93 160.9 178.9 42 135 28 132.9 158.9 10 119.1 6 - -

2a 1-CQAL 334.9 160.9 172.9 98 288.8 25 133 ------

5d 4-FQA 366.94 172.8 190.8 20 192.8 11 93.2 111.1 75 136.9 27 71.5 26

165

Supplementary table 5. Chlorogenic acids in Brazil cold brew coffee sample 15 MS 1 MS 2 MS 3

No Compd. Parent base peak Secondary peak base peak secondary peak ion m/z m/z int m/z int m/z m/z int m/z int m/z int

4c 3-CQA 352.93 190.8 178.9 41 176.9 17 126.9 85.3 58 172.8 49 93.3 31

6c M-3-CQ 366.94 160.8 192.8 16 134.9 9 132.9 133.9 6 117.1 4 - -

4e 5-CQA 352.9 190.8 214.8 13 191.8 8 394.7 126.9 40 93.2 31 172.9 24

5c 3-FQA 366.99 192.9 134 15 193.8 14 133.9 148.9 16 136.4 3 117 2

4d 4-CQA 352.93 172.8 178.8 43 190.9 21 93.2 111.1 60 154.9 46 94.2 26

2c 3-CQAL 334.93 160.9 135 84 161.9 13 132.9 ------

6d M-4-CQ 366.97 160.9 134.9 32 161.8 12 132.9 ------

1e 5-CSA 334.91 172.9 178.8 16 317.8 9 111.1 93.2 71 155 20 59.7 14

5e 5-FQA 366.98 190.8 172.9 5 348.9 5 85.3 172.8 74 111.1 54 126.9 41

2d 4-CQAL 334.89 160.9 178.8 59 210.8 37 132.9 143.9 25 - - - -

5d 4-FQA 366.96 172.9 347.9 18 192.8 11 93.2 110.9 52 71.5 48 154.9 23

166

3.4. Research on Application to Other Food

Identification of Composition Kinetics of Traditional Korean Green Tea

Producing Method by LC-MS

Seung-Hun Lee 1, Inamullah Hakeem Said 1, Anastasiia Shevchuk 1, Un-Jae Lee 2, Nikolai

Kuhnert 1*

Department of Life Sciences and Chemistry, Jacobs University Bremen, 28759 Bremen,

Germany 1

Keumro Handmade Tea, Jeollabuk-do, Korea 2

*Correspondence to:

Prof. Dr. Nikolai Kuhnert,

Department of Life Sciences and Chemistry

Jacobs University Bremen gGmbH

Campus Ring 1

28759 Bremen, Germany

Tel: +49 421 200 3120

Fax: +49 421 200 3102

E-mail: [email protected]

167

Abstract

Nine times steaming and drying process is originated by making Korean Red Ginseng and is adapted to green tea parching process. The samples are prepared in a traditional way Korean tea making process separating by each step. All the time series samples are analyzed by HPLC-TOFMS and HPLC-IontrapMS for interpreting and quantifying polyphenols and caffeine. Those samples were analyzed both all the components in the leaves and the content of tea infusions consumed. It could be shown that repeated steaming increases the extractability of catechins.

Keywords: Korean traditional green tea; Nine times steaming and drying; Polyphenol;

Korean tea ceremony

168

1. Introduction

The Korean traditional green tea method did not attract scientific interest since its production scale is much lower than machine producing methods. However, in March

2016, Korean Cultural Heritage Administration reported that Korean traditional green tea producing method would become Korean national intangible cultural heritage in

2016. 1 Therefore, to identify the extracts of the tea produced by the Korean traditional method merits scientific interest, based on its importance in cultural heritage.

Nine repetitive parching and drying process, which is one of the Korean traditional green tea production methods, has been adapted from nine repetitive steaming and drying process used in Korean red ginseng production. In the Korean red ginseng method, an increase of health promoting saponins during repetition of the parching process has already been observed. 2 This nine times repeating method is known to the public to increase the value of the product.

Starting from the explanation for French Paradox, 3 polyphenols attracted scientific interest. Scientists found polyphenols from wine, fruits, coffee and tea and concatenated them to the scientific grounds of healthy food. Green tea constitute one of the richest source of flavanols within the human diet with abundant amount of (epi)catechin (EC), gallocatechin (GC), (epi)gallocatechin gallate (EGCG), (epi)catechin gallate (ECG) and so on. 4

In 2003, three different regioisomers of caffeoylquinic acids were separated and analyzed with HPLC-tandem MS. 5 Influenced by this research, not only for the

169

chlorogenic acids, but also the whole polyphenol compounds could be easily analyzed by checking their fragmentation. After this research, HPLC-tandem MS became the most power tool for analyzing all the different families of polyphenols. 6-8 Especially for thearubigins, HPLC-iontrap MS and UPLC-ionmobility MS were selected and developed to analyze compounds in black tea. 9,10

In this research, we analyzed sequential samples by 9 times repetitive parching and drying of green tea leaves with two different sampling methods; extracting with organic solvent by sonication of finely ground tea leaves, and aqueous brewing following the traditional Korean tea ceremony. The two different extracted teas were analyzed by LC- high resolution ESI-MS and LC-tandem ESI-MS to identify and quantify polyphenols and caffeine. EIC spectra of specific polyphenols were used to analyze the content changes by sequential parching samples.

2. Experimental

The experimental methods of HPLC and LC/MS n are followed by previous publications of our group research. 11-13

2.1. Chemicals and Green Tea Samples

All the chemicals (Analytical grade) were purchased from Sigma-Aldrich (Germany).

Green tea samples in different parching processes were provided from Geumro

Handmade Tea in Jeollabuk-do, Korea.

2.2. Kinetic parching samples or green tea

170

Parching for green tea leaves were conducted as to make commercial samples. Starting from sample 0 as without parching, green tea leaves were processed with continuous heating and drying. In the standard nine-times heating and drying process, the first parching temperature starts from higher than 250 ℃, gradually decreases to 50 ℃ for the ninth process. Every sample was numbered and took out from the pot to make samples from number 1 to 9. All the samples were sealed separately and kept in room temperature until use.

2.3. High Resolution MS

The LC equipment (Agilent 1100 series, Karlsruhe, Germany) comprised a binary pump, an autosampler with a 100 µL loop, and a diode-array detector (DAD) with a light-pipe flow cell (recording at 320 and 254 nm and scanning from 200 to 600 nm). This was interfaced with a MicroTOF Focus mass spectrometer (Bruker Daltonics, Bremen,

2.4. LC/MS n

171

collision energy. Maximum fragmentation amplitude was set to 1 V, starting at 20 % and ending at 200 %. MS operating conditions (negative mode) had been optimized using 5-caffeoylquinic acid with a capillary temperature of 365 ℃, a drying gas flow rate of 10 L / min, and a nebulizer pressure of 10 psi.

2.5. HPLC

Separation was achieved on a 150 × 3 mm i.d. column containing diphenyl 5 µm, with a

5 mm × 3 mm i.d. guard column (Varian, Darmstadt, Germany). Solvent A was water / formic acid (1000:0.005 v/v) and solvent B was methanol. Solvents were delivered at a total flow rate of 500 µL / min. The gradient profile was from 10 to 80 % B linearly in

70 min followed by 10 min a return to 10 % B and 10 min isocratic to re-equilibrate.

2.6. Green tea extracting

2.6.1. Total polyphenol extracting

The sample preparation method for total polyphenol extracting used ultra-sonication for obtaining sufficient extracting yield following the previous published methods for analyzing polyphenols of plant leaves. 14,15 In this research, 2 g of dried green tea leaves for each sample and put them in a mortar to grind into powder with the temperature of liquid nitrogen. Falcon tubes of 15 mL were used to mix 1 g of the powder into 10 mL of 70 % methanol solution. The tubes were ultra-sonicated for 30 min and filtered through a membrane filter. The solutions were used directly for LC-MS.

2.6.2. Practical brewing polyphenol extracting

172

The sample preparation method for practical polyphenol extracting was tried to follow to the Korean traditional green tea ceremony. 1 g of green tea sample was put into a 250 mL beaker, followed by 100 mL deionized water filling. After 2 min of waiting without stirring, the extracted tea solution was carefully poured out and filtered through a membrane filter. The filling water, waiting and pouring out extracted solution were repeated twice to make three different sequential brewing samples. Those filtered solutions were directly used for LC-MS.

3. Results and Discussion

3.1. Interpretations of polyphenol

MicroTOF was used to obtain exact molecular weight to determine chemical formula and Iontrap MS was used to get the fragmentation patterns to determine isomeric structures of polyphenol compounds in Korean green tea samples.

3.1.1. Molecular formula analysis of high resolution MS data

Polyphenolic compounds and caffeine peaks are detected and listed with their Sodium adduct in positive mode by TOF MS in Table 1 . The experimental value is based on the

1st water extracted non-parching sample. After obtaining the chemical formula by high resolution MS data, tandem MS data were used together to determine the isomeric composition of each compound.

173

Table 1 . High resolution mass data in the 1 st water extracted samples of non-parching

Korean green tea sample in positive ion mode

Exp. m/z Theor. m/z No. RT Name Mol. Formula Error (ppm) (M+H) (M+H)

1 6.1 Quinic acid C7H13 O6 193.0696 193.0707 5.4 + 2 6.1 Na adduct of Quinic acid C7H12 NaO 6 215.0523 215.0526 1.5

3 25.2 Gallocatechin C15 H15 O7 307.0816 307.0812 -1.2 + 4 25.2 Na adduct of Gallocatechin C15 H14 NaO 7 329.0631 329.0632 0.3

5 25.3 3-CQA C16 H19 O9 355.1054 355.1024 -8.6 + 6 25.3 Na adduct of 3-CQA C16 H18 NaO 9 377.0813 377.0843 7.9

7 27.5 procyanidin B2 C30 H27 O12 579.1464 579.1497 5.7 Na + adduct of procyanidin 8 27.5 C H NaO 601.1272 601.1316 7.3 B2 30 26 12 + 9 29.4 Na adduct of 5-CQA C16 H18 NaO 9 377.0845 377.0843 -0.5

10 30.4 Caffeine C8H11 N4O2 195.087 195.0877 3.6 + 11 30.4 Na adduct of Caffeine C8H10 N4NaO 2 217.0689 217.0696 3.3

12 30.4 3-CoQA C16 H19 O8 339.1083 339.1074 -2.5 + 13 30.4 Na adduct of 3-CoQA C16 H18 NaO 8 361.0885 361.0894 2.4

14 32 (epi)gallocatechin gallate C22 H19 O11 459.0914 459.0922 1.7 Na + adduct of 15 32 C H NaO 481.0724 481.0741 3.5 (epi)gallocatechin gallate 22 18 11

16 32.1 cis-3-CoQA C16 H19 O8 339.1071 339.1074 1 + 17 32.1 Na adduct of cis-3-CoQA C16 H18 NaO 8 361.089 361.0894 1.2

18 33.6 (epi)catechin C15 H15 O6 291.0879 291.0863 -5.5 + 19 33.6 Na adduct of (epi)catechin C15 H14 NaO 6 313.0693 313.0683 -3.3

20 36.8 5-CoQA C16 H19 O8 339.1042 339.1074 9.6 + 21 36.8 Na adduct of 5-CoQA C16 H18 NaO 8 361.0879 361.0894 4.1

22 39.2 (epi)catechin gallate C22 H19 O10 443.0958 443.0973 3.3 Na + adduct of (epi)catechin 23 39.2 C H NaO 465.078 465.0792 2.7 gallate 22 18 10

24 39.6 4-CoQA C16H19 O8 339.107 339.1074 1.3 + 25 39.6 Na adduct of 4-CoQA C16 H18 NaO 8 361.0883 361.0894 2.9

26 41.1 cis-4-CoQA C16 H19 O8 339.1072 339.1074 0.7 + 27 41.1 Na adduct of cis-4-CoQA C16 H18 NaO 8 361.0892 361.0894 0.4

28 41.2 p-Coumaroylquinic acid C19 H19 O8 339.1075 339.1074 -0.1 Na + adduct of p- 29 41.2 C H NaO 361.0889 361.0894 1.5 Coumaroylquinic acid 19 18 8

30 56.1 Theaflavin-3-O-gallate C36 H29 O16 717.1378 717.145 10.1 Na + adduct of Theaflavin-3- 31 56.1 C H NaO 739.1256 739.127 1.8 O-gallate 38 28 16

174

3.2.2. Compound assignment

Compounds were identified based on fragmentation data from tandem-MS spectra. The

methanolic soxhlet extract of non-parching Korean green tea sample spectrum was used

to assign isomeric structures of polyphenols. Table 2 shows assignment of polyphenols

with their fragmentation patterns from the base peak chromatogram shown in Figure 1 .

Table 2. Tandem MS data in negative ion mode of polyphenols from non-parching

Korean green tea methanolic soxhlet extract

MS 1 MS 2 MS 3

No RT Compd. Parent base peak Secondary peak base peak secondary peak ion m/z m/z int m/z int m/z m/z int m/z int m/z int

1 4.1 Caffeoylhexose dimer 683.27 341 341.9 10 340.2 1 178.9 160.9 38 113.1 23 143 20

2 5.9 QA dimer 383.06 190.8 191.8 8 347 1 126.9 93.2 61 172.9 52 85.3 34

3 6 QA 190.88 127 85.4 54 93.3 45 109.1 0 0 0 0 0 0

4 11.2 Gallic acid 168.87 125 123 83 0 0 95.2 0 0 0 0 0 0

5 22.7 3-CQA 352.98 190.8 178.8 41 176.8 13 85.3 97.2 49 87.2 36 109 32

6 24.3 Gallocatechin 304.91 178.9 220.9 80 218.9 65 163.9 135 51 150.9 39 137 24

7 25.4 Catechin 288.92 244.8 204.8 36 178.8 11 203 161 30 226.8 24 186.8 22

8 27.1 Procyanidin B2 577.17 407 425 79 289 27 284.8 280.8 87 389 49 242.8 30

9 28.3 5-CQA 352.98 190.8 214.8 13 178.8 5 127 111 31 85.4 30 172.8 29

10 28.7 3-CoQA 336.96 162.9 119.1 10 163.8 9 119 120 4 117 0 0 0

11 30.5 cis-3-CoQA 336.96 162.8 190.8 8 119 7 119 0 0 0 0 0 0

12 31.5 (epi)Gallocatechin gallate 457.05 168.9 330.9 29 304.9 24 125 123.1 4 97.4 3 107 0

13 32.6 (epi)Catechin 288.91 244.9 204.8 42 202.9 15 202.9 186.9 17 160.9 17 226.9 16

14 35.3 5-CoQA 336.96 190.8 163 4 0 0 127 85.4 67 172.8 53 93.2 37

15 37.9 4-CoQA (de) 336.93 172.8 163 6 0 0 93.2 111 61 154.8 20 71.6 19

16 38.1 (epi)Catechin gallate 441.05 288.9 330.9 25 289.9 21 244.9 204.8 40 202.9 14 178.9 13

17 39.6 4-CoQA 336.94 172.9 173.8 8 190.9 6 93.2 111.1 49 71.5 22 154.9 17

18 55.4 Theaflavin-3-O-gallate 715.15 527 563.2 66 545 63 371 399 69 483.2 46 509 23

175

Figure 1. Base peak chromatogram in negative ion mode of non-parching Korean

Green Tea methanolic soxhlet extract

3.2. Results on different number of times on parching

3.2.1. Comparison on linear fit of different extracting methods of the sequential parching samples

The UV Integrations of selected compounds’ peaks in HPLC spectra of Korean Green

Tea samples were compared in Figure 2 . Composition changes by number of parching of methanolic extracted samples are shown in Figure 2(a) and their linear fits are shown in Figure 2(b) . Three times averaged composition changes by number of parching of water extracted samples are shown in Figure 2(c) with their linear fits in Figure 2(d) .

176

18000 18000 (b) (a) 15000 15000

12000 GC Caffeine 12000 GC EC Caffeine

9000 ECG EC

9000 EGCG ECG EGCG 6000 UV Intensity (mAU) Intensity UV

UV Intensity (mAU) Intensity UV 6000

3000 3000

0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9

50000 40000

35000 40000 GC Caffeine 30000 EC 30000 GC ECG (c) Caffeine 25000 EGCG EC

20000 ECG 20000 EGCG (d) 15000 UV Intensity (mAU) Intensity UV 10000 (mAU) Intensity UV 10000

0 5000

0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9

Figure 2. Composition changes of the sequential parching samples and their linear fits;

(a) composition of methanolic extracts (b) linear fit of methanolic extracts (c) composition of water extracts (d) linear fit of water extracts

Notable differences between the two different extracting solvent data are the different slopes of linear fits shown in Figure 2(b) and Figure 2(d) . In methanolic extracted samples, the parching times does not affect to the extraction yields of each compound as presented in Figure 2(b) . On the other hand, the more parching times, the more yields of polyphenolic compounds are obtained on water extracted samples, presented in

Figure 2(d) . This phenomenon could be explained by the following assumption that the

9 times parching and dry process does not decompose the flavonoids in tea, so the total

177

phenolics are not changing, but this process can soften the green tea leaves to obtain better extraction yields. The detailed variables of linear fits on each compound in two different solvent extracted data are shown in Supporting Information Figure 1 to 10 .

3.2.1. Comparison on multiple brewing samples following Korean traditional tea ceremony.

The traditional Korean tea ceremony recommends three times repeating tea brewing for one batch of tea leaves. As it is indicated in 2.6.2. Practical brewing polyphenol extracting, the experiment followed traditional method of tea brewing. The entire sequential composition spectra are shown in Figure 3 classified by the compounds.

178

a) b) 30000 First First Second 75000 Second Third Third 24000 Average Average

60000 18000

12000 45000 UV Intensity (mAU) Intensity UV UV Intensity (mAU) Intensity UV 6000 30000

0 15000 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9

75000 c) d) First First Second 16000 Second 60000 Third Third Average Average

12000 45000

30000 8000 UV Intensity (mAU) UV Intensity 15000 (mAU) UV Intensity 4000

0 0 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9

e) 25000 First Second Third 20000 Average

15000

10000 UV Intensity (mAU) Intensity UV

5000

0 0 1 2 3 4 5 6 7 8 9

Figure 3. Composition of the sequential parching samples extracted by practical brewing method; a) Gallocatechin b) Caffeine c) (Epi)Gallocatechin gallate d)

(Epi)Catechin e) (Epi)Catechin Gallate

179

As shown in Figure 3, tea flavonoids’ concentrations start to decrease in the third brewing. In traditional Korean tea ceremony, dried tea leaves start to be unrolled from the first brewing. Therefore, the first brewing does not have enough surface area to be extracted. This is the reason why the first and the second brewings do not have meaningful flavonoid concentration disparity. The third brewing seems the starting point of diminution. After considering these phenomena, the Korean tea ceremony could be regarded as a scientific consuming method of green tea flavonoids.

4. Conclusion

In conclusion, nine times parching and drying green tea samples were analyzed and interpreted with two different extracting methods; soxhlet extraction with aqueous methanol for total polyphenol analysis and practical brewing method with water for sequential sample analysis. LC-TOFMS and LC-IontrapMS were used for analyzing the extracts to interpret polyphenols and caffeine in each sample.

The most important discovery in this study is the different polyphenol contents between total polyphenol content and actual consuming polyphenol contents. Although the total polyphenol contents were not changed by its different parching steps, actual consuming polyphenol contents increased in accordance with the parching step proceeds. As a result, the nine times parching method is a soft method which does not decompose the polyphenols but powerful method which influences the extracting yield by practical tea brewing. On top of that, by analyzing the sequential brewing concentration, Korean

180

green tea ceremony is verified as a scientifically sound tea consuming method through three times brewing repetition for efficient flavonoids extraction.

The nine times heating and drying method was mainly investigated only for the Korean red ginseng process and started to spread throughout other foods. The discovery in this study is important by itself for the green tea development and is expected for adapting into the process development of the other food innovation.

Acknowledgements

Excellent technical support by Anja Müller is acknowledged.

References

1. Korean Intangible Heritage Digital Archive Korean Traditional Tea Process. 2017,

2. Kim, D.; Kim, Y.; Lee, Y.; Min, J.; Kim, S.; Yang, D. Conversion of ginsenosides by 9 repetitive steamings and dryings process of Korean ginseng root and its inhibition of BACE-1 activity. Journal of Physiology & Pathology in Korean Medicine 2008, 22,

3. Renaud, S.; Delorgeril, M. Wine, Alcohol, Platelets, and the French Paradox for Coronary Heart-Disease. Lancet 1992, 339, 1523-1526.

4. Graham, H.N. Green tea composition, consumption, and polyphenol chemistry. Preventive Medicine 1992, 21, 334-350.

5. Clifford, M.N.; Johnston, K.L.; Knight, S.; Kuhnert, N. Hierarchical scheme for LC- MSn identification of chlorogenic acids. J. Agric. Food Chem. 2003, 51, 2900-2911.

181

6. Jaiswal, R.; Kuhnert, N. Hierarchical scheme for liquid chromatography/multi-stage spectrometric identification of 3,4,5-triacyl chlorogenic acids in green Robusta coffee beans. Rapid Communications in Mass Spectrometry 2010, 24, 2283-2294.

7. Jaiswal, R.; Kuhnert, N. Identification and characterization of five new classes of chlorogenic acids in burdock (Arctium lappa L.) roots by liquid chromatography/tandem mass spectrometry. Food & Function 2011, 2, 63-71.

8. Kuhnert, N.; Jaiswal, R.; Matei, M.F.; Sovdat, T.; Deshpande, S. How to distinguish between feruloyl quinic acids and isoferuloyl quinic acids by liquid chromatography/tandem mass spectrometry. Rapid Communications in Mass Spectrometry 2010, 24, 1575-1582.

9. Yassin GH, Grun C, Koek JH, Assaf KI, Kuhnert N. Investigation of isomeric flavanol structures in black tea thearubigins using ultraperformance liquid chromatography coupled to hybrid quadrupole/ion mobility/time of flight mass spectrometry. J Mass Spectrom. 2014;49(11):1086-1095.

10. Yassin GH, Koek JH, Kuhnert N. Model system-based mechanistic studies of black tea thearubigin formation. Food Chem. 2015;180:272-279.

13. Kuhnert, N.; Jaiswal, R.; Eravuchira, P.; El-Abassy, R.M.; von der Kammer, B.; Materny, A. Scope and limitations of principal component analysis of high resolution LC-TOF-MS data: the analysis of the chlorogenic acid fraction in green coffee beans as a case study. Analytical Methods 2011, 3, 144-155.

182

14. Saito, S.T.; Gosmann, G.; Saffi, J.; Presser, M.; Richter, M.F.; Bergold, A.M. Characterization of the constituents and antioxidant activity of Brazilian green tea (Camellia sinensis var. assamica IAC-259 cultivar) extracts. J. Agric. Food Chem. 2007, 55, 9409-9414.

15. Sereshti, H.; Samadi, S.; Jalali-Heravi, M. Determination of volatile components of green, black, oolong and white tea by optimized ultrasound-assisted extraction- dispersive liquid-liquid microextraction coupled with gas chromatography. J. Chromatogr. A 2013, 1280, 1-8.

183

18000 Gallocatechin, MeOH (aq) Linear fit of Gallocatechin 15000

12000

9000

UV Intensity (mAU) Intensity UV 6000

3000

0 1 2 3 4 5 6 7 8 9

Y = 4342.72512 + 71.59505 * X R = 0.22088 P = 0.53972 Supporting Information Figure 1. Linear fit of Gallocatechin extracted by 70 % methanol (aq)

184

18000 Caffeine, MeOH (aq) Linear fit of Caffeine 15000

12000

9000

UV Intensity (mAU) Intensity UV 6000

3000

0 1 2 3 4 5 6 7 8 9

Y = 4069.66854 – 57.85839 * X R = -0.43275 P = 0.21161 Supporting Information Figure 2. Linear fit of Caffeine extracted by 70 % methanol (aq)

185

18000 Epicatechin, MeOH (aq) Linear fit of Epicatechin 15000

12000

9000

UV Intensity (mAU) Intensity UV 6000

3000

0 1 2 3 4 5 6 7 8 9

Y = 2196.08093 + 14.03555 * X R = 0.09681 P = 0.79019 Supporting Information Figure 3. Linear fit of Epicatechin extracted by 70 % methanol (aq)

186

18000 Epicatechin Gallate, MeOH (aq) Linear fit of Epicatechin Gallate 15000

12000

9000

UV Intensity (mAU) Intensity UV 6000

3000

0 1 2 3 4 5 6 7 8 9

Y = 4416.79979 + 25.24202 * X R = 0.10403 P = 0.77487 Supporting Information Figure 4. Linear fit of Epicatechin gallate extracted by 70 % methanol (aq)

187

18000

15000

12000

9000 Epigallocatechin Gallate, MeOH (aq) Linear fit of Epigallocatechin Gallate

UV Intensity (mAU) Intensity UV 6000

3000

0 1 2 3 4 5 6 7 8 9

Y = 13067.91881 + 167.27783 * X R = 0.17709 P = 0.62454 Supporting Information Figure 5. Linear fit of Epigallocatechin gallate extracted by 70 % methanol (aq)

188

50000 Gallocatechin, Water Linear fit of Gallocatechin 40000

30000

20000 UV Intensity (mAU) Intensity UV 10000

0 1 2 3 4 5 6 7 8 9

Y = 3277.95293 + 1184.37219 * X R = 0.83607 P = 0.00258 Supporting Information Figure 6. Linear fit of Gallocatechin extracted by Water

189

50000

40000

30000

20000 Caffeine, Water Linear fit of Caffeine UV Intensity (mAU) Intensity UV 10000

0 1 2 3 4 5 6 7 8 9

Y = 37981.13732 + 213.50734 * X R = 0.14388 P = 0.6917 Supporting Information Figure 7. Linear fit of Caffeine extracted by Water

190

50000 (Epi)Catechin, Water Linear fit of (Epi)Catechin 40000

30000

20000 UV Intensity (mAU) Intensity UV 10000

0 1 2 3 4 5 6 7 8 9

Y = 3143.54382 + 658.77461 * X R = 0.93223 P < 0.0001 Supporting Information Figure 8. Linear fit of Epicatechin extracted by Water

191

50000 (Epi)Catechin Gallate, Water Linear fit of (Epi)Catechin Gallate 40000

30000

20000 UV Intensity (mAU) Intensity UV 10000

0 1 2 3 4 5 6 7 8 9

Y = 3920.76593 + 1165.60438 * X R = 0.93467 P < 0.0001 Supporting Information Figure 9. Linear fit of Epicatechin gallate extracted by Water

192

50000 (Epi)Gallocatechin Gallate, Water Linear fit of (Epi)Gallocatechin Gallate 40000

30000

20000

UV Intensity (mAU) Intensity UV 10000

0 1 2 3 4 5 6 7 8 9

Y = 7692.49752 + 3242.9969 * X R = 0.84168 P = 0.00226 Supporting Information Figure 10. Linear fit of Epigallocatechin gallate extracted by Water

193

4. Conclusion

In this study, the hierarchical key LC-MS n method was mainly used for identifying CGAs and caffeine of different experimental samples from different stages of the coffee processes; harvesting and grading, roasting, brewing and applications.

By these researches, below are the first results ever reported with scientific methodology.

1. The chlorogenic acids of Jamaica Blue Mountain coffee were investigated.

2. The different grade green coffee beans in a single origin were compared.

3. The possibility as quality influencing factor of ratios of caffeoylquinic acids’

regioisomers were suggested.

4. Higher antioxidant property processed coffee bean than green coffee bean was found.

5. The differences of aqueous and organic solvent extracted coffee were proved.

6. Cold brew coffee, which is the recent brewing method, was analyzed.

7. The composition changes by kinetic samples of coffee were identified.

8. A new caffeine control method other than decaffeination was developed.

9. The applicability of coffee analyzing methods to other food techniques was confirmed.

In the first harvesting and grading step, three different grades of Jamaica Blue Mountain coffees were compared each other and with other originated coffees. By using LC-MS n method, 47 different CGAs and their derivatives were identified as other published coffee papers reported, but first time investigated for the Jamaica Blue Mountain coffee beans and different grade coffees in the same origins. Additionally, the regioisomer ratio of 3-, 4- and 5-

CQA was suggested as the possible aroma factor for the first time as the specialty Jamaica beans and other normal beans had different values.

194

In the roasting process, novel processed coffee beans were investigated to search the possibility as roasting process substitute. The novel process was nine times repeating of parching and drying method which is derived from Korean traditional red-ginseng process so this special method is named steamentation from steaming and fermentation. Most importantly, the antioxidant property of the novel method was higher than any other processed coffee bean samples including green coffee bean and two different degree roasted coffee beans. Especially, higher antioxidant property processed coffee bean than the green coffee bean was first time reported in this study. EIC at m/z 367 and 335 were concentrically interpreted by LC-MS n method and showed the differences of chemical reactivity of steamented coffee to green and roasted coffees.

In the brewing stage, maximum 14 times different 5-CQA equivalent was found with 3-CQA methyl ester integration in 70 % aqueous methanol and water extracted samples. This result demonstrated that a discrepancy between polyphenol levels determined from optimized extraction protocols and levels found in actual coffee brews prepared by consumers.

Moreover, cold-brew coffee, the recently became popular coffee brewing method coffee, was analyzed separated in sequential order separation. The most important discovery in this research is that the caffeine in the same solid content increases in time dependent manners of the samples. It means that only separating the coffee while brewing can be a way of caffeine control method controlling ± 25 % difference if compared to normal coffee consumption. This is the only caffeine control method other than the decaffeination, and it costs nothing from normal coffee brewing method unlike the decaffeination method. Therefore, we named this innovative method LeeN-Caffeination from less- and enhanced- caffeine coffee making method developed by Seung-Hun Lee and Nikolai Kuhnert.

195

Traditional Korean green tea samples were selected to expand the applicability of the techniques obtained from the steps of coffee process. The selected samples were regarded as specialty tea like Jamaica Blue Mountain in coffee, employed the same Korean red-ginseng process like steamented coffee, used different extraction method for comparing lab research and practical consuming like different extracting solvent and the brewing was separated three times like LeeN-Caffeination. The most important discovery in this study was that the polyphenols are increasing by repeating the parching method in practical brewing, though the total polyphenols in the green tea leaves were not changing. This means that the special parching method might be soft not to destroy the polyphenols in the green tea, but it is strong enough to weaken the tea leaves to extract polyphenols easily in practical brewing. Also it is proven that all the methods used in coffee in this research might be adaptable to other food process or analysis.

By conducting all the projects in different coffee processes and application, it is found that the

LC-MS n method could be useful for comparing different developments on different coffee processes. The continuous searching of new developments of coffee products and adapting the identification methods are important for the coffee consumer’s health.

The coffee market has been increasing and a lot of new processes have been invented in all different processes of the coffee processing steps. A new fermentation control methods, sprout coffee, roasting machines only using electricity, transparent coffee, Nitrogen brewing and all sort of new coffee products are developed and consumed without careful identification.

There could be special composition changes by controlling the coffee processes and some composition changes can have critical effect for specific diseases; caffeine for high blood pressure, chlorogenic acids for type II diabetes or acidity for dental health. The result of this research is expected to be used for the new developments of coffee in order to give correct 196

information so that all coffee consumers can select coffee according to their individual health conditions.

197