Food Research International 119 (2019) 683–692

Contents lists available at ScienceDirect

Food Research International

journal homepage: www.elsevier.com/locate/foodres

NMR analysis of roasted coffee lipids and development of a spent ground T

coffee application for the production of bioplastic precursors
⁎
Kathryn Williamsona, Emmanuel Hatzakisa,b,
a
Department of Food Science and Technology, The Ohio State University, Parker Building, 2015 Fyffe Rd, Columbus, OH, United States
b
Foods for Health Discovery Theme, The Ohio State University, United States

A R T I C LE I N FO A B S T R A C T

Keywords: Multinuclear and multidimensional NMR spectroscopy was applied as a robust and rapid tool for the analysis of
NMR several classes of non-polar compounds in roasted coffee beans, coffee beverage and spent coffee grounds. In
Coffee lipids addition to various fatty acids, other compounds found in roasted coffee lipids, include oxidation and hydrolysis
scCO2 extraction products, terpenes, sterols, and phospholipids. Spent coffee grounds have a similar fatty acid composition with
Bioplastic precursors
roasted coffee beans and they are rich in Cafestol and Kawheol, which appear as esters of fatty acids.
Triglycerides extracted from coffee waste using a green chemistry approach, based on supercritical CO2 ex-
traction, are promising candidates for the production of bioplastics. Bioplastic precursors were produced using
an in situ solvent-free epoxidation process and the reaction monitoring was performed using NMR spectroscopy.

1. Introduction industries. Due to its large number of current and potential applications
in the food, cosmetic, pharmaceutical and bioenergy industry, coffee oil
Coffee is among the most traded and consumed commodities in the is valued at a higher price compared to coffee itself. It can be used as a
world. The two most commercially available coffee species are Coffea sun protectant because of its ability to block UV-B radiation, as well as
arabica and Coffea canephora, commonly referred to as Arabica and help with the formation of connective tissue (Velazquez et al., 2009)
Robusta coffee. The mass of a green Arabica coffee bean is composed of and help with skin dryness and serious dermatological conditions, such
about 15–17% lipids, while a green Robusta bean is composed of about as eczema and psoriasis (Raba et al., 2015). Another diverse application
10% lipids (Calligaris, Munari, Arrighetti, & Barba, 2009). These lipids, of coffee oil is using the oil obtained from the wasted spent coffee
collectively called coffee oil, are composed of nearly 75% triacylgly- grounds (SCG) following coffee beverage extraction as a source for
cerols (TG) (Calligaris et al., 2009; Speer & Kölling-Speer, 2006), but biodiesel production (Oliveira, Franca, Camargos, & Ferraz, 2008; Park,
unlike many oils of plant or animal origin, coffee oil also contains a Kim, & Lee, 2016) and for the microbial-based production of biode-
unique unsaponifiable fraction (Calligaris et al., 2009). gradable biopolymers, such as polyhydroxylalkanoates (PHA) and poly
Because of its fatty acid (FA) composition and unique profile of (3-hydroxybutyrate) (PHB) (Campos-Vega, Loarca-Pina, Vergara-
bioactive compounds, coffee oil has many applications as a health Castaneda, & Oomah, 2015). Oil and carbohydrates obtained from SGC
promoting, natural product. Two coffee-specific diterpenes, Cafestol have been used for the production of value-added products and thus
and Kahweol, are particularly important for health (George, render SCG an attractive target for valorization (Kovalcik, Obruca, &
Ramalakshmi, & Rao, 2008) and they are also used as biomarkers for Marova, 2018).
coffee processing and coffee authentication (Scharnhop & Winterhalter, Coffee oil has been analyzed using a variety of analytical methods
2009). Due to its peculiar taste and flavoring, coffee oil is used as a including High Performance Liquid Chromatography (HPLC), Gas
basic ingredient for the production of many food products, such as Chromatography (GC), Capillary gas chromatography (cGC),
candies, cakes, and beverages, where it is added to not only improve the Electrospray ionization-mass spectrometry (EIS-MS), and Fourier
sensory characteristic in a product, but also increase the health pro- Transform Infrared Spectroscopy (FT-IR) (Valdenebro, León-Camacho,
moting effects in foods (Raba et al., 2015). Pablos, Gonzalez, & Martin, 1999; Carrera, Leo, Pablos, & Gonza Âlez,
In addition to its nutritional properties, coffee oil has the potential 1998; González, Pablos, Martín, Leon-Camacho, & Valdenebro, 2001;
in an array of applications in the cosmetic and pharmaceutical Martín, Pablos, González, Valdenebro, & Leon-Camacho, 2001). Despite

⁎
Corresponding author at: Department of Food Science and Technology, The Ohio State University, Parker Building, 2015 Fyffe Rd, Columbus, OH, United States.
E-mail addresses: [email protected] (K. Williamson), [email protected] (E. Hatzakis).

https://doi.org/10.1016/j.foodres.2018.10.046
Received 22 August 2018; Received in revised form 9 October 2018; Accepted 12 October 2018
Available online 15 October 2018
0963-9969/ © 2018 Elsevier Ltd. All rights reserved.
K. Williamson, E. Hatzakis Food Research International 119 (2019) 683–692

the benefits of these methods, some of them require lengthy extraction Cambridge Isotope Laboratories (Tewksbury, MA). Cr(acac)3, 3,5-Di-
techniques, derivatization, and complex sample preparation, or may tert-butyl-4-hydroxytoluene (BHT), hydrogen peroxide, formic acid,
not be capable of providing quantitative data or information on the NaCl, NaHCO3, EDTA-K+, chloroform, petroleum ether, ethyl acetate,
molecular level. For example, the GC analysis of fatty acids requires and ethanol were obtained from Fisher Chemical (Waltham, MA, USA).
separation and derivatization of individual compounds, during which,
lipid oxidation may occur and can cause further inaccuracies (Guillén & 2.3. Production of coffee oil epoxides
Ruiz, 2003; Igarashi et al., 2000; Sacchi, Medina, Aubourg, Addeo, &
Paolillo, 1993). Nuclear Magnetic Resonance (NMR) Spectroscopy is a 2.3.1. scCO2 extraction
robust method that can rapidly analyze the lipid profile of coffee on the 160 g of spent coffee waste was lyoplilized (FreeZone Plus 12,
molecular level without requiring separation and/or purification steps. Kansas City, MO, USA) and placed in a supercritical fluid extraction
The analysis is conducted in one snapshot, which provides a significant system (Waters Inc., Milford, MA, USA) with 1 L sample capacity.
advantage compared to traditional analysis that focus on one specific Coffee oil was extracted using 100% CO2 as extraction medium (70 °C,
class of compounds. The major drawback of NMR is its low sensitivity 300 bar, 100 g/min). The process consisted of five cycles in total. Each
compared to other analytical techniques, such as chromatography and cycle (20 min) contained one stage of dynamic extraction (10 min)
mass spectrometry, however NMR's nondestructive and rapid, robust followed by one stage of static extraction (10 min). The total run time
nature offers a great alternative. was 100 min.
NMR has been used in the past to analyze polar coffee components
for classification purposes, to differentiate Arabica and Robusta coffee, 2.3.2. SCG oil epoxidation
to analyze green coffee according to variety and origin, and to measure Coffee oil was epoxidized at a molar ratio of 1.0:1.0:5.6 (coffee oil:
changes that occur during roasting (Ciampa, Renzi, Taglienti, Sequi, & formic acid: hydrogen peroxide). First, 2.5 g of the extracted oil was
Valentini, 2010; Monakhova et al., 2015; Schievano, Finotello, De mixed with hydrogen peroxide (1.794 g, 30%) in a round bottom flask
Angelis, Mammi, & Navarini, 2014; Wei et al., 2012). So far, NMR has and stirred at 400 rpm to form a homogenous mixture. Next, 0.5 mL of
been used by D'Amelio et al. to evaluate green coffee lipids of coffee formic acid (0.154 g, 85%) was added drop-wise and the mixture was
beans from different origins and species (Arabica and Robusta) stirred at 50 °C for 12 h. Epoxidized oil was taken from the reaction
(D'Amelio, De Angelis, Navarini, Schievano, & Mammi, 2013). In that mixture and transferred into a separatory funnel where it was mixed
study, the results obtained by NMR were compared to those obtained by with 8.3 mL of ethyl acetate. After shaking the funnel, the mixture was
GC and found to be in very good agreement. In addition, 1H NMR left to relax until the two layers were separated. The water (lower) layer
spectroscopy has been recently applied to determine the presence of was removed and the organic layer was shaken with 12.5 mL brine
Arabica and Robusta species by screening lipophilic extracts of coffee (saturated salt solution). The mixture was left to relax until the two
beans (Finotello et al., 2017; Monakhova et al., 2015). In general, NMR layers separated, and the water layer was removed. 12.5 mL of a
has been extensively validated for the quantitative determination of NaHCO3 1 M solution was added and shaken to neutralize the excess of
various lipid components from different sources and is now considered acid in the solution. The water phase was discarded and the organic
as a reliable method for this type of analysis (Dais et al., 2007; D'Amelio phase was further washed with 4.5 mL brine. Finally, the organic phase
et al., 2013; Igarashi et al., 2000; Knothe & Kenar, 2004; Sacchi et al., was collected and dried using 3.5 g of sodium sulfate (anhydrous), fil-
1993). tered (4 Whatman), and the solvent was removed under reduced
Here, we apply multinuclear (1H, 13C, 31P) and multidimensional pressure evaporation (Buchi, Flawil, CH).
NMR spectroscopy for the determination of various coffee lipid com-
ponents in roasted coffee beans, as well as SCG and the coffee beverage. 2.4. Sample preparation for NMR experiments
Further, we investigated the potential of coffee oil extracted from SCG,
a waste product with limited value, for the production of lipid epoxides, 2.4.1. Coffee oil from green and roasted beans
which are precursors for bioplastics. Bioplastics produced from vege- 2 g of coffee was ground using a L'equip mini coffee grinder & seed
table oils have excellent mechanical and physicochemical properties. mill (Nutrimill, Salt Lake City, UT) for 30 s. Ground samples were
Thus, finding renewable and low cost alternative sources to partially mechanically shaken with 10 mL of chloroform for 5 min at room
replace petrochemical-based polymers is an attractive and sustainable temperature. The extracts were filtered using filter paper (4 Whatman)
approach (Montero De Espinosa & Meier, 2011). For the coffee oil ex- and the solvent was removed under reduced pressure evaporation. The
traction, we adopted a green chemistry approach using supercritical oil was dissolved in 600 μL of a stock solution (20 mL) composed of
CO2 (scCO2). Epoxides were produced using formic acid:hydrogen CDCl3/DMSO d6 in 5:1 volume ratio containing 10 mg of chromium
peroxide as an oxidizing agent in a solvent-free system, and were acetylacetonate, Cr(acac)3 (28.6 uM), and 110 mg of BHT (500 uM),
characterized based on previous literature and multidimensional NMR protected from moisture with 5A molecular sieves.
spectroscopy. Reaction monitoring and quantification was performed
using the 1D spectra. 2.4.2. Coffee oil extracted from coffee beverage
The coffee beverage from the roasted coffee beans was prepared at
2. Materials and methods 5% coffee: water. 20 mL of brewed coffee was shaken with 20 mL of
CHCl3 or petroleum ether. Each mixture was centrifuged at 2700 rpm
2.1. Samples for 20 min at 20 °C (Eppendorf Centrifuge 5810R, Hamburg, DE). The
bottom CHCl3 layer and the upper petroleum ether phase was collected
Ten roasted coffee beans samples, five from Africa and five from in a rotary evaporator flask. The remaining coffee was extracted a
South America, all processed under the same roasting conditions, were second time with an additional 15 mL of CHCl3 or petroleum ether
used in this study. All samples were harvested in 2015 and stored at under the same centrifugal parameters. The solvent from the two ex-
−40 °C under the same conditions. Spent coffee grounds were collected tractions was combined and removed under reduced pressure eva-
after brewing from a local coffee shop in Columbus, OH. poration. The remaining oil was dissolved in 600 μL of the same
CDCl3:DMSO (5:1) mixture as mentioned above and placed into a 5 mm
2.2. Chemicals NMR tube.

Chloroform d1 (CDCl3), Dimethyl sulfoxide-d6 (DMSO), deuterated 2.4.3. Phospholipids extraction from ground coffee
water (D2O), and methanol-d (CD3OD), were purchased from 5 g of ground coffee were extracted by 2 × 20 mL of ethanol/water

684
K. Williamson, E. Hatzakis Food Research International 119 (2019) 683–692

(85:15 v/v). The ethanolic extracts were washed with 10 mL of petro- determination of compounds bearing labile protons, such as hydro-
leum ether, and the phospholipids were collected upon removal of the peroxides with OOH protons, in the 1H NMR spectra. This solvent
solvent under vacuum. The residual oil that contained the phospholi- mixture has been found to yield spectra with excellent resolution
pids was re-dissolved in 0.5 mL of CDCl3/CD3OD/D2O-EDTA-K+ (Skiera, Panagiotis, Kuballa, Holzgrabe, & Diehl, 2012) as DMSO d6
(400:95:5 v/v/v). forms hydrogen bonds with OOH protons (Merkx, Hong, Ermacore,
Duynhoven, & M., 2018) and slows down the proton exchange.
2.4.4. Coffee oil and epoxides from coffee waste-reaction monitoring
2.5 mL of sample was taken after 5 h and after 10 h of reaction time 3.1.1. Fatty acids
and was treated as described in Section 2.3. After the removal of ethyl The total amount of unsaturated and saturated fatty acids (SFA),
acetate the epoxidized oil was dissolved in 0.5 mL of CDCl3. The solu- except Linolenic (LN), can be determined by the signal that appears at δ
tion was then transferred into a 5 mm NMR tube. 0.88 in the 1H NMR spectrum and belongs to the protons of the terminal
methyl group of fatty acids (Fig. 1A). For most oils with a plant or
2.5. NMR experiments animal origin, these methyl protons appear as triplets. However, in case
of coffee oil they appear as a pseudo-quartet that arises from the
1
H and 13C NMR experiments were conducted on a Brüker Avance overlapping between the methyl protons of oleic acid (OL) and those of
III spectrometer (Bruker, Ettlingen, DE) operating at 700.13 MHz and linoleic acid (LO), which appear in increased amounts in coffee oil.
176.04 MHz for 1H and 13C nuclei, respectively, equipped with a TXO Acquiring the spectrum in instruments that operate in higher Larmor
helium-cooled 5 mm probe. 31P NMR experiments were carried out on a frequencies (eg 850 MHz) reveals two distinct triplets (Fig. S1). A
Brüker Avance III spectrometer operating at 323.89 MHz for 31P nuclei, convenient way for the determination of individual FA, such as OL and
equipped with a TCI helium-cooled 5 mm probe. Certain 2D NMR ex- LO using a 1H NMR spectrum is the use of the distinct signals of their
periments were also performed on a spectrometer operating at allylic protons, at δ 2.00 and δ 2.04, respectively. LO can also be de-
850.23 MHz and 213.81 MHz for 1H and 13C nuclei, respectively, termined by the bis-allylic protons at δ 2.76. The signal of the bis-allylic
equipped with a triple resonance helium-cooled inverse (TCI) 5 mm protons at δ 2.79 is recommended for determination of LN, instead of
probe. All experiments were performed at 25 ± 0.1 °C and the spectra the signal of allylic protons because the LN allylic protons overlap with
were processed by the Topspin software package provided by Brüker the protons H1 and H14 of Cafestol as found by the HSQC-TOCSY
Biospin. spectrum. The CH2(α) protons of FA acyl chains at δ 2.37–2.22 can be
used as an internal reference that represents the total amount of FA, for
2.5.1. NMR spectra relative quantifications of different FA using the following equations:
1
H NMR spectra were recorded using the following acquisition
ILO I I
parameters: 16 scans and 4 dummy scans, 64 K data points, 90° pulse CLO = , COL = OL , CLN = LN , CSFA = IS − CLO − COL − CLN
IS 2IS 2IS
angle (10.5 μs), relaxation delay 10 s to ensure quantitative results,
spectral width 13 ppm. A polynomial fourth-order function was applied where CLO, COL, CLN, CSFA, are the relative concentrations of LO, OL, LN
for base-line correction in order to achieve accurate quantitative mea- and SFA respectively, and ILO, IOL, ILN and IS are the integral of the
surements upon integration of signals of interest. The spectra were signals at δ 2.76, δ 2.00, δ 2.79 and δ 2.37–2.22, respectively.
acquired without spinning the NMR tube in order to avoid artifacts, The 13C NMR spectrum can also be very informative for the analysis
such as spinning side bands of the first or higher order. Chemical shifts of fatty acids that appear in coffee oil. LO can be determined from
are reported in ppm from TMS (δ = 0). carbon C9 at δ 129.25 or the bis-allylic carbon at δ 23.93, OL from C9 at
13
C NMR spectra were obtained with proton decoupling using the δ 128.73 (sn-1,3) and δ 128.70 (sn-2) and LN from carbon C15 at δ
inverse gated decoupling pulse sequence to minimize NOE effects. 126.23 or carbon C16 at δ 131.89 (Zamora, Alba, & Hidalgo, 2001). For
Repetition delays between consecutive 90° pulses are equal to five times 13
C analysis the carbon C3 at δ 26.22 was used as a reference for re-
the longitudinal relaxation times measured by the null method. 13C lative quantification. No normalization is required for carbon analysis
NMR spectra were recorded with spectral width of 200 ppm, using 64 K and proportions of FA result from the ratio between the diagnostic
data points, a 90° excitation pulse (9.8 μs); acquisition time 0.9 s and signal and the signal of C2.
relaxation delay of 60 s in order to avoid signal saturation. 128 scans Table 1 summarizes the quantitative data, expressed in relative
were collected and spectra zero-filled to 128 K. For all FIDs, line concentrations for the FA found in coffee oil extracted from roasted
broadening of 1 Hz was applied prior to Fourier transform. Chemical beans. Five samples of Arabica beans from Africa and five samples of
shifts are reported in ppm from CDCl3 (δ = 77). Arabica beans from South America, roasted under the same conditions,
Experimental details and pertinent references for most of the 2D were used and no correlation between the geographical origin and the
pulse sequences used in this study can be found in elsewhere (Berger & FA composition was observed. A systematic deviation for OL was found
Braun, 2004; Dais, Misiak, & Hatzakis, 2015). between 1H and 13C NMR analysis. This is probably because that in
contrast to other oils, coffee oil contains minor compounds that bear
3. Results and discussion allylic protons in their structures and thus have signals that appear at
the same chemical shifts as those of OL. The fatty acid composition of
3.1. Compound identification-spectral assignment in roasted coffee coffee oil as determined in this study is similar to other studies (Raba
et al., 2015; Somnuk, Eawlex, & Prateepchaikul, 2017), however some
Both 1H and 13C NMR spectra contain diagnostic signals that can be differences may be found due to the extraction method, the origin of
attributed to specific compounds which can be used for their quantifi- coffee bean and the type/degree of roasting.
cation. Because the reliability of the qualitative and quantitative ana-
lysis depends on the correct NMR assignment of the 1D spectra, we 3.1.2. Diterpenes, sterols and caffeine
performed the first, to our knowledge, high resolution systematic 2D NMR offers an efficient, alternative approach for the analysis of
NMR analysis that provides the unambiguous assignment of 1D 1H and Cafestol and Kahweol in coffee oil, which is less time consuming than
13
C NMR signals of various coffee oil constituents. traditional chromatographic separation. The two coffee specific di-
Fig. 1A and B display the high resolution 1H NMR spectrum and the terpenes are structurally very similar which can cause challenges in
olefinic region of the 13C NMR spectrum of coffee oil obtained from a their separation required for chromatographic analysis. (Carrera et al.,
roasted coffee bean, acquired in CDCl3:DMSO d6 (5:1) solution. The use 1998). The molecular formulas and numbering systems of Cafestol and
of a solvent mixture that contains DMSO-d6 is essential for the Kawheol are shown in Scheme SI. The signal which is most convenient

685
K. Williamson, E. Hatzakis Food Research International 119 (2019) 683–692

Fig. 1. (A) 700 MHz 1H NMR spectrum and (B) 176 MHz 13
C NMR spectrum of coffee oil extracted from a roasted coffee bean, acquired in CDCl3:DMSO d6 (5:1)
solution.

686
K. Williamson, E. Hatzakis Food Research International 119 (2019) 683–692

Table 1
Fatty acids composition (%) of ten Arabica coffee oil samples from Africa and South America as determined by 1H and 13
C NMR spectroscopy.
Compound, diagnostic signal 1 2 3 4 5 6 7 8 9 10

1
H NMR
LO (2.76, 2H, t, J = 7.25 Hz) 39.90 43.67 41.88 42.98 41.70 43.65 44.85 41.85 45.01 41.66
OL (4H, 2.00, m) 9.68 9.47 10.21 9.56 11.12 9.23 9.13 11.01 8.91 10.18
LN (2.79, 4H, t, J = 6.29 Hz) 1.91 1.40 1.58 1.51 1.52 1.51 1.38 1.73 1.38 1.51
SFA 48.51 45.46 46.33 45.95 45.66 45.61 44.64 45.41 44.7 46.65
13
C NMR
LO (129.16) 41.82 42.04 43.41 41.97 41.13 41.97 41.99 42.65 43.92 40.02
OL (128.75) 7.06 7.65 9.03 7.67 8.27 7.16 7.88 9.11 7.19 8.54
LN (13.90) 1.24 1.31 1.15 1.42 1.53 1.51 1.30 1.49 1.32 1.21
SFA 49.88 49.00 46.41 48.94 49.07 49.36 48.83 46.75 47.57 50.23

LO, Linoleic acid; OL, Oleic acid; LN, Linolenic acid; SFA, Saturated Fatty Acids.
1, Ethiopia; 2, Kenya; 3, Rawanda; 4, Kenya, 5, Uganda; 6, Colombia; 7, Guatemala; 8, Colombia; 9, Guatemala; 10, Colombia.

for the determination of Cafestol appears at δ 7.22 and belongs to

proton H19, whereas Kahweol can be easily determined by its signal at
δ 7.27, attributed to proton H19. Kahweol has an extra AB system with
two olefinic protons H2 and H1, which appear as doublets at δ 6.21 and
δ 5.89, respectively, with a J coupling of 10.1 Hz. These signals can also
be used for Kahweol's quantification. The corresponding C19 carbons of
Cafestol at δ 139.57 and Kahweol at δ 140.20 can be easily determined
by the HSQC-DEPT spectrum as shown in Fig. S2. The quaternary car-
bons C3 and C4 can be identified from the HMBC spectrum. C3 and C4
of Cafestol appear at δ 147.74 and δ 119.30 respectively and C3, C4 of
Kawheol appear at δ 149.27, 120.94 respectively. These assignments
are in agreement with previous literature (Scharnhop & Winterhalter,
2009), although significant deviations were observed due to the dif-
ference in the solvent mixture between NMR experiments. The NMR
analysis also confirmed the chemical shifts of other protons, such as the
methyl protons H20 of Cafestol and Kaweol at δ 0.82 and δ 0.96, re-
spectively. The corresponding carbons at δ 12.42 and δ 14.69 can be
assigned from the HSQC-DEPT spectrum. The diastereotopic protons
H17a and H17b of Kawheol and Cafestol that form an AB system, ap-
pear at δ 4.28 and δ 4.21, respectively with a J coupling of 11.4 Hz. The Fig. 2. The band selective constant time HMBC spectrum of a roasted coffee
corresponding carbons C17 of Kawheol and Cafestol have a signal at δ lipids sample acquired over a 4 ppm spectral width, in CDCl3:DMSO d6 (5:1)
67.40. NMR indicates that Cafestol and Kawheol appear in the form of solution.
esters of fatty acids, rather than the free form, which is in agreement
with previous studies (Scharnhop & Winterhalter, 2009). The signals at of scans and thus increased experimental time.
δ 173.15 and δ 173.12 belong to the carboxyl esters group of these Caffeine in coffee oil can be easily determined using the 1H NMR
compounds, as found by the correlation peak in the HMBC spectra with signals of the aromatic proton H2 at δ 7.66 and the methyl protons H10,
H17 protons (Fig. 2). H14 and H12, which all appear as singlets at δ 3.98, 3.55 and δ 3.36,
The main sterols that appear in coffee are β-Sitosterol, respectively. Quantifications can also be performed using the signals of
Stigmatasterol, and Campesterol (Guercia, Berti, Navarini, Demitri, & carbons C2, C10, and C12 that resonate at δ 141.09, 32.61, 26.84, re-
Forzato, 2016), which belong to the group of 4-desmethylsterols. 4- spectively, as determined by the HSQC-DEPT experiment. The carbon
desmethylsterol consumption has been shown to exhibit antitumor, C14 at δ 28.71 cannot be used for quantification purposes because of its
antidiabetic and antifungal properties (Abdul, Choi, Jung, & Choi, overlapping with C4 carbons of fatty acids. For absolute quantifications,
2016), as well as cause decrease the blood cholesterol levels (de Jong, of sterols, caffeine and terpenes in coffee oil, BHT was used as an in-
Plat, & Mensink, 2003). In addition, they have been found to be very ternal standard, as described previously (Williamson & Hatzakis, 2017).
efficient tools for coffee authentication purposes (Carrera et al., 1998). Table S1 shows quantitative data for Cafestol, Kawheol, Stigmatasterol
These three sterols represent more than 90% of the sterolic fraction of and Caffeine for ten roasted samples from Africa and South America.
coffee oil and can be reliably determined by 1H NMR (Hatzakis, While more samples are needed to make definite conclusion, the pre-
Dagounakis, Agiomyrgianaki, & Dais, 2010). β-Sitosterol and Campes- sent data show no correlation between these compounds and the geo-
terol both resonate at δ 0.68 because of the close isochronicity of their graphical origin. As also mentioned for fatty acid composition, these
methyl groups at position 18, while the methyl protons at C18 of results are comparable to those obtained by previous studies (Belandria
Stigmatasterol have a signal at δ 0.70 which can be used for its quan- et al., 2016; Kamm et al., 2002), however some deviations may be
tification. In instruments operating at Larmor frequencies higher than observed because factors, such as roasting, geographical origin and
700 MHz, β-Sitosterol and Campesterol have distinct signals at δ 0.679 harvesting year can affect the biochemical profile of a coffee bean. In
and δ 0.682, which allow for their individual determination. The signal addition, a single extraction with chloroform was used in this study, in
at δ 0.54 is accredited to the C18 methyl group of Δ7-stigmastenol and contrast to Soxhlet extraction which is applied when the total amount
Δ7-avenasterol which are known to appear in coffee oil and resonate in of coffee lipids is desired. This also explains why the amounts of Caf-
that chemical shift (Speer & Kölling-Speer, 2006; Wilson et al., 1996). feine measured in this study are significantly lower than those reported
13
C NMR can also be used for the analysis of β-Sitosterol, Stigmatasterol in literature, as Caffeine is a relatively polar compound and only small
and Campesterol from the signals of C18 at δ 10.95, δ 11.08 and δ amounts are extracted with chloroform. This means that all of the
11.14, respectively, however this analysis requires an increased number

687
K. Williamson, E. Hatzakis Food Research International 119 (2019) 683–692

Caffeine that can be found in the coffee bean itself may not be present protons in the 1H NMR spectrum at δ 4.09 and 2′-CHOH at δ 4.00.
in the oil extract. Carbons C1′ and C3′ of 1,3-DG give one signal at δ 64.13 which can be
used for quantitative purposes. The signal of the C2′ of 1,3-DGs appears
3.1.3. Oxidation-hydrolysis products and phospholipids at δ 66.0 as revealed by the HSQC-TOCSY and HSQC-DEPT spectra. The
The lipid fraction of both green and roasted coffee beans contains carboxylic carbons of 1,3-DG at position sn-1,3 appear at δ 172.49, as
primary and secondary oxidation products, such as hydroperoxides and found by their correlation peak with 1, 3-CH2OCO protons in the se-
aldehydes (Bosco, Toffaninm, Palo, Zatti, & Segre, 1999), as a result of lective HMBC spectrum (Fig. 2). 1,2-DGs appear only in small amounts
the oxidation of unsaturated fatty acids. These compounds play an in coffee and can be determined by their signals at δ 5.07 (2′a-CHOCO),
important role in the sensory perception of a final coffee beverage and and δ 3.65 (3′a, 3′b-CH2OCO). The signals at δ 4.34 (1′b-CH2OCO) and
can potentially be used as indicators for the oxidative changes that δ 4.14 (1′α-CH2OCO) cannot be used for quantification of 1,2-DGs
occur in coffee during roasting or storage. However, they are also because of overlapping. However, the carbons C1′, C2′ and C3′ resonate
known to have genotoxicity and cytotoxicity (Guillén & Goicoechea, at δ 61.68, δ 71.16, and δ 59.53 and can all be used for quantification,
2008). The 1H NMR spectrum is a promising tool for future kinetic although a significantly increased number of scans is required for
studies as it can display changes that take place at the molecular level, achieving a reliable S/N.
even in three weeks of storage of ground coffee (Fig. S3). It is worth Phopsholipids have been found in coffee, and the current metho-
noting that for the same storage period, no changes were observed in dology for their determination is based on Folch method followed by
the corresponding whole coffee beans, indicating that grinding favors HPLC analysis (Zhou et al., 2013). Because of the small amounts of
lipid oxidation. phospholipids that appear in coffee, the 1H and 13C NMR signals of
The main secondary oxidation product in coffee oil has a signal at δ phospholipids usually overlap with the signals of the major lipids in
10.12 (dd, J = 7.01 Hz), which is characteristic for aldehydic protons. coffee and cannot be used for their determination. 31P NMR is a pro-
The application of a window function for resolution enhancement re- mising alternative for their analysis in various food products (Burri,
veals a four-bond coupling of 1.65 Hz. This proton has a cross peak in Hoem, Monakhova, & Diehl, 2016; Kaffarnik, Ehlers, Gröbner,
the HSQC-DEPT spectrum with a methine carbon at δ 191.83 and be- Schleucher, & Vetter, 2013). The challenge with this 31P analysis is that
longs to the same spin system with the olefinic protons at δ 5.94 (bd, PLs form aggregates and electrostatic complexes with paramagnetic
J = 10.10 Hz) as found by the HSQC-TOCSY spectrum. In addition, ions, which has a negative impact in the resolution and sensitivity of the
there are correlation peaks in HSQC-TOCSY between the protons at δ 31
P NMR spectra. The first problem can be confronted using a solvent
5.94 and δ 10.12 with the protons at δ 6.44 (1H, ddd, J = 15.19 Hz, mixture of CDCl3/CD3OD/D2O-EDTA-K+ (400:95:5 v/v/v) and the
10.67 Hz, 3.75 Hz) and δ 5.64 (1H, dt, J = 15.19 Hz, J = 7.01). Two latter by washing the phospholipids solution with EDTA-cation salts
allylic protons at δ 2.11 and three methylic protons at δ 0.93 were also (Hatzakis, Koidis, Boskou, & Dais, 2008). Fig. 3 shows the 1H–31P-
found to belong to the same molecule, however we were not able to HMBC spectrum of phospholipids extracted from coffee oil. The che-
identify the corresponding carbons because they overlap with those of mical shifts order correspond to Phosphatidylcholine (PC), Lyso-
FA. The multiplicity pattern of the signals indicates the presence of an Phosphatidylcholine (LPC) Phosphatidyethanolamine (PE) and Phos-
α,β-unsaturated aldehyde. Close inspection of spectral properties, such phatidylinositol (PI).
as chemical shifts, J coupling values, and relative integral ratios led us
to attribute these assignments to an E, Z 2,4-dienal.
3.2. Lipids in coffee beverage
The other main signal in the aldehyde region appears as a broad
singlet at δ 9.68. This can be assigned to a quaternary aldehyde due to
Small amounts of lipids can be found in the water extracts of coffee,
its multiplicity and the absence of any correlation peak in the HSQC-
however, they form aggregates with short T2 relaxation times and thus
TOCSY spectrum. Utilization of signal integrals and coupling constants
they appear as a broad peak at δ 0.92, which belongs to the terminal
helped us associate the singlet at δ 9.68 with the signals at δ 7.16 and δ
methyl protons of fatty acids (Fig. S4). For this reason, they cannot be
6.70, which belong to two aromatic protons that form an AB system and
studied without extraction from the beverage using a non-polar solvent,
have a cross peak in the TOCSY spectrum, and the aliphatic protons at δ
such as chloroform or petroleum ether. The dominant NMR signals in
4.55. This chemical assignment had been initially attributed to 5-hy-
chloroform and the petroleum ether extracts of a coffee beverage be-
droxymethyl-2-furaldehyde, which has been found in coffee previously,
long to caffeine. However, both extracts are also rich in various lipid or
however this scenario was not confirmed because of the lack of com-
lipid-related components, as shown in Fig. 4. The spectral area from δ
mercially available standard until recently (Bosco et al., 1999). Spiking
13.0–10.0 (Fig. 4A) is rich in enols, lipid peroxides and carboxylic
of coffee lipids sample with 5-hydroxymethyl-2-furaldehyde, revealed
that although it exists in small amounts in the samples, the signal at δ
9.68 does not belong to it so the most prominent candidate is an aro-
matic aldehyde, as has been also proposed by Bosco et al. In some
samples a broad triplet that overlaps with the singlet at δ 9.68, and
belong to n-alkanals (Guillen & Goicoechea, 2009), also appears.
The main hydrolysis products in coffee oil are 1,3-, 1,2-diglycerides
(DG) and free fatty acids (FFA). Some small amounts of MG may also
appear in some samples. The band selective constant time HMBC
spectrum optimized for three-bond CeH coupling provides increased
spectral resolution in the indirect dimension (F1), compared to the
regular HMBC. This selective HMBC spectrum revealed the chemical
shift assignments of the carbonyl carbon signals of FA in the form of
1,3-DG and the signals Hα protons of FFA, shown in Fig. 2. The total
amount of FFA can be determined by the triplet at δ 2.24 (J = 6.44 Hz)
in the 1H NMR spectrum, which belongs to the CH2(α) protons of FFA
and has a correlation peak in the band-selective constant time HMBC
spectrum. The carboxylic carbons at δ 174.77 and δ 174.74 can also be
used for FFA determination. Fig. 3. 800 MHz 1H-31P HMBC spectrum of phospholipids isolated from roasted
1,3-DG can be determined by the distinct signals of 1,3-CH2OCO coffee grounds, in CDCl3/CD3OD/D2O-EDTA-K+ (400:95:5 v/v/v).

688
K. Williamson, E. Hatzakis Food Research International 119 (2019) 683–692

acids, and several aldehydes appear in the δ 10.0–9.0 region (Fig. 4B). 3.3. Extraction and epoxidation of coffee waste lipids
The spectra contain several signals attributed to aromatic compounds
such as phenols and their derivatives (Ruiz-Aracama, Goicoechea, & For the production of coffee beverage, roasted coffee grounds are
Guillén, 2017), as well as oxygenated compounds with double bonds, extracted with water, a polar solvent, leaving behind a large amount of
such as hydroxy- and hydroperoxy-conjugated-dienic systems, as can be the non-polar lipids in the SCG to be wasted. This coffee waste is a rich
seen in the area from δ 9.0 to δ 5.90 (Fig. 4C). Finally the typical signals source of triglycerides with an excellent quality and a relatively large
of methylene and methyl protons of FA at δ 1.26 and δ 0.88, as well as unsaturated profile. NMR analysis also shows that SCG have a FA
small amounts of sterols at δ 0.68, can also be observed (Fig. 4D). In profile very similar to roasted coffee (Fig. 5A). Lipids, such as TG and
contrast to findings from Bosco et al., small amounts of diglycerides FA, have been previously used for the production of biopolymers that
were found in the coffee brews used in this study (Fig. 4C). Interest- can substitute polymers derived by petrochemicals (Montero De
ingly, only small amounts of Cafestol and Kaweol were found in the Espinosa & Meier, 2011; Plaza et al., 2017). Therefore, SCG utilization
coffee brew, which is in agreement with a recent study (Rendón et al., for the production of resins and biopolymers is an attractive approach
2017) and indicates that a majority of the Cafestol and Kaweol remains for achieving a low cost, renewable application. One possible pathway
in the SCG. This result is confirmed by the 1H NMR spectrum of lipids for creating bioplastics from raw plant material lipids is epoxidation for
isolated from SCG in the aromatic region (Fig. S5), which indicates the the creation of epoxy resins (Abdullah & Salimon, 2010). Epoxy resins
presence of Kaweol and Cafestol in large amounts, which renders SCG can be created through the introduction of an easily polymerizable and
as a rich and economic source of these valuable compounds. A similar more reactive functional group to a TG, such as an epoxide. TG are, in
trend was also found for sterols. Although some traces of sterols were general, neutral molecules with low reactivity, so these TG become
found in the beverage (Fig. 4A), most remained in the SCG (Fig. S5). epoxy resins with much more applicability. Epoxy resins contain one or
While further analysis combining NMR spectroscopy and/or other more epoxide groups and following a series of chemical reactions, they
analytical methods such as gas chromatography (GC) is required to become polymers with strong mechanical properties and high tem-
obtain more comprehensive information for individual substances, the perature and chemical resistance.
current study provides an NMR screening that was able to identify the For creation of epoxy resins, SCG were first freeze-dried to increase
various classes of compounds. The NMR spectra of lipids isolated from the yield of the extraction. Because the oil from SCG contains significant
coffee beverage can be used for metabolomics studies to identify pat- amounts of oxidation and hydrolysis products that may affect the re-
terns between lipid composition and coffee brewing properties. action and the stability of the products, oil was initially extracted using
scCO2, which is a very non-polar extraction medium that favors the

Fig. 4. Expansions of a 700 MHz 1H NMR spectrum of lipids extracted from coffee brew in CDCl3:DMSO d6 (5:1) solution showing the regions of peroxides/enols/FFA
(A), aldehydes (B), aromatic compounds and oxygenated compounds with double bonds (C) and methyl protons of lipids (D).

689
K. Williamson, E. Hatzakis Food Research International 119 (2019) 683–692

Fig. 5. Epoxidation process as monitored by 1H NMR spectroscopy. (A) Before epoxidation, (B) after 5 h of reaction, (C) after 10 h of reaction, (D) HSQC-DEPT
spectrum of epoxidation products.

extraction of non-polar lipids, such as TG and FFA, while semi-polar ethanol to the process, while making the process more complex, allows
compounds such as lipids peroxides, MG and DG are extracted at lower for improved extraction and allows the process to remain green.
yields. An additional advantage of this supercritical fluid technology However, the addition of solvents such as hexane or chloroform will no
technique is that it does not require the use of organic solvents and is a longer allow for a green extraction process. The quality and chemical
green replacement to Sohxlet extraction. The average extraction yield composition of the oil extracted by spent coffee grounds was monitored
of the scCO2 extraction was ~ 13% (w/w), which is comparable to the using 1H and 13C NMR spectroscopy and was found to be similar to the
extraction yield reported in other studies (Akgün, Bulut, Kikic, & coffee oil extracted from roasted beans (LO, 45–50%; SFA, 38–45; OL,
Solinas, 2014), however it is important to note that various factors such 8–11%; LN, 1–3%). The FA composition is affected by the method of
as variety, growing year, altitude, roasting, etc. may cause a coffee extraction (scCO2 or organic solvent) as shown in a previous study
sample to have varying amounts of oil or extractable oil. The yield can (Ahangari & Sargolzaei, 2013). The extracted oil was then epoxidized
be improved by the addition of organic solvents. The addition of in-situ using H2O2 and formic acid and the reaction was monitored by

690
K. Williamson, E. Hatzakis Food Research International 119 (2019) 683–692

NMR as previously described for canola oil (Xia, Budge, & Lumsden, Hatch project 1012460. Authors wish to thank Dr. Rafael Jiménez-
2016). In the epoxidation reaction, H2O2 initially reacts with formic Flores for providing technical assistance and access to scCO2 extraction
acid to form peroxy-formic acid and water. In a next step an epoxy device and Dr. Devin Peterson for providing the coffee samples. This
group is added on the double bonds of the unsaturated fatty acid chains. study made use of the Campus Chemical Instrument Center NMR fa-
The reaction scheme and reaction monitoring using NMR is shown in cility at Ohio State University.
Fig. 5.
It was observed that the SCG lipids have remarkable stability to Conflict of interest
hydrolysis during storage. Even after 3 weeks of storage at room tem-
perature, no increase in the hydrolysis products was observed in non- The authors declare no conflict of interest.
dried SCG. This is probably because SCG are a product of roasted coffee,
in which the hydrolytic enzymes such as lipases have been deactivated References
during roasting. Coffee oil is free of epoxides before the reaction,
whereas two possible isomers are possible after reaction initiation, as Abdul, Q. A., Choi, R. J., Jung, H. A., & Choi, J. S. (2016). Health benefit of fucosterol
shown if Fig. 5C. This is supported by the fact that four distinct signals from marine algae: A review. Journal of the Science of Food and Agriculture, 96,
1856–1866.
of the epoxidized coffee oil are observed at δ 3.30–2.90, as a result of Abdullah, B. M., & Salimon, J. (2010). Epoxidation of vegetable oils and fatty acids:
the insertion of epoxide groups in TG molecules. An efficient and clear Catalysts, methods and advantages. Journal of Applied Sciences, 10, 1545–1553.
indicator for the yield and the status of the reaction are the signals Ahangari, B., & Sargolzaei, J. (2013). Extraction of lipids from spent coffee grounds using
organic solvents and supercritical carbon dioxide. Journal of Food Processing and
intensities of olefinic protons at δ 5.42–5.29, which are reduced as the Preservation, 37, 1014–1021.
reaction takes place. Small amounts of monoepoxides are still in the Akgün, N. A., Bulut, H., Kikic, I., & Solinas, D. (2014). Extraction behavior of lipids ob-
mixture even after the completion of the reaction, as found by the tained from spent coffee grounds using supercritical carbon dioxide. Chemical
Engineering & Technology, 37, 1975–1981.
signals at δ 2.92 and the olefinic signals at δ 5.57–5.39, which belong to Belandria, V., Oliveira, P. M. A., Chartier, A., Rabi, J. A., Oliveira, A. L., & Bostyn, S.
olefinic protons located next to an epoxide group. Signals that were (2016). Pressurized-fluid extraction of cafestol and kahweol diterpenes from green
observed between δ 8.5–7.9 belong to peroxides and other oxirane ring- coffee. Innovative Food Science and Emerging Technologies, 37, 145–152.
Berger, S., & Braun, S. (2004). 200 and more experiments: A practical course. Weinheim:
opening reaction side-products, due to the presence of acid in the so-
Wiley VCH.
lution (Milchert, Smagowitcz, & Lewandowski, 2010). These side-pro- Bosco, M., Toffaninm, R., Palo, D., Zatti, L., & Segre, A. (1999). High-resolution 1H NMR
ducts were only formed in very minor concentrations and don't affect investigation of coffee. Journal of the Science of Food and Agriculture, 79, 69–878.
the overall yield of the process. It is also worth noting that the chemical Burri, L., Hoem, N., Monakhova, Y. B., & Diehl, B. W. (2016). Fingerprinting krill oil by
31
P, 1H and 13C NMR spectroscopies. Journal of the American Oil Chemists' Society, 93,
shift of methyl groups of LO and OL are not significantly affected by the 1037–1049.
epoxidation, but the methyl group of LN is affected because there is a Calligaris, S., Munari, M., Arrighetti, G., & Barba, L. (2009). Insights into the physico-
double bond in n-3 position that is epoxidized. The 13C NMR spectrum chemical properties of coffee oil. European Journal of Lipid Science and Technology,
111, 1270–1277.
can also be used for reaction monitoring and determination of epoxides. Campos-Vega, R., Loarca-Pina, G., Vergara-Castaneda, H., & Oomah, B. D. (2015). Spent
Fig. 5D displays the HSQC-TOCSY spectrum of the epoxidized coffee oil coffee grounds: A review on current research and future prospects. Trends in Food
extracted from SCG. These epoxides can be used for the production of Science & Technology, 45, 24–36.
Carrera, F., Leo, Â.-C., Pablos, F., & Gonza Âlez, A. G. (1998). Authentication of green
resins and biopolymers by utilizing several different approaches coffee varieties according to their sterolic profile. Analytica Chimica Acta, 370,
(Montero De Espinosa & Meier, 2011). For the quantification of ep- 131–139.
oxides and the determination of the reaction yield, the signal of sn-2 Ciampa, A., Renzi, G., Taglienti, A., Sequi, P., & Valentini, M. (2010). Studies on coffee
roasting process by means of nuclear magnetic resonance spectroscopy. Journal of
glyceridic protons at δ 5.23, which are not affected by the epoxidation Food Quality, 33, 199–211.
reaction, were used. Results indicated that in less than 10 h, which is Dais, P., Misiak, M., & Hatzakis, E. (2015). Analysis of marine dietary supplements using
much shorter than canola oil epoxidation, more than 95% of coffee oil NMR spectroscopy. Analytical Methods, 7, 5226–5238.
Dais, P., Spyros, A., Christoforidou, S., Hatzakis, E., Fragaki, G., Agiomirgianaki, A., ...
FA have been converted to epoxides. Further increasing the reaction
Brenes, M. (2007). Comparison of analytical methodologies based on 1H and 31P
time does not significantly affect the yield of the reaction. NMR spectroscopy with conventional methods of analysis for the determination of
some olive oil constituents. Journal of Agricultural and Food Chemistry, 55, 577–584.
4. Conclusions D'Amelio, N., De Angelis, E., Navarini, L., Schievano, E., & Mammi, S. (2013). Green
coffee oil analysis by high-resolution nuclear magnetic resonance spectroscopy.
Talanta, 110, 118–127.
This study showed that multinuclear and multidimensional NMR Finotello, C., Forzato, C., Gasparini, A., Mammi, S., Navarini, L., & Schievano, E. (2017).
spectroscopy is a powerful tool for the high resolution analysis of lipids NMR quantification of 16-O-methylcafestol and kahweol in Coffea canephora var.
robusta beans from different geographical origins. Food Control, 75, 62–69.
in roasted coffee beans, coffee beverage and SCG. Several classes of George, S. E., Ramalakshmi, K., & Rao, L. J. M. (2008). A perception on health benefits of
compounds can be determined from their diagnostic signals as assigned coffee. Critical Reviews in Food Science and Nutrition, 48, 464–486.
by various 2D NMR experiments, such as HSQC-DEPT, band-selective González, A. G., Pablos, F., Martín, M. J., Leon-Camacho, M., & Valdenebro, M. S. (2001).
HPLC analysis of tocopherols and triglycerides in coffee and their use as authenti-
constant-time HMBC, HSQC-TOCSY, etc. SCG, a waste product with cation parameters. Food Chemistry, 73, 93–101.
limited alternate uses, have a similar FA composition with roasted Guercia, E., Berti, F., Navarini, L., Demitri, N., & Forzato, C. (2016). Isolation and
coffee beans and they are a rich source of valuable lipids, which can be characterization of major diterpenes from C. canephora roasted coffee oil.
Tetrahedron: Asymmetry, 27, 649–656.
efficiently extracted using scCO2 as an alternative to Sohxlet extraction
Guillén, M. D., & Goicoechea, E. (2008). Toxic oxygenated α,β-unsaturated aldehydes and
that utilizes organic solvents. TGs extracted from SCG were successfully their study in foods: A review. Critical Reviews in Food Science and Nutrition, 48,
converted in-situ to epoxides using H2O2 and formic acid and thus they 119–136.
Guillen, M. D., & Goicoechea, E. (2009). Oxidation of corn oil at room temperature:
are promising candidates for the production of bioplastics. Reaction
Primary and secondary oxidation products and determination of their concentration
monitoring and product characterization was performed using NMR in the oil liquid matrix from 1H nuclear magnetic resonance data. Food Chemistry,
spectroscopy. 116, 183–192.
Supplementary data to this article can be found online at https:// Guillén, M. D., & Ruiz, A. (2003). 1H nuclear magnetic resonance as a fast tool for de-
termining the composition of acyl chains in acylglycerol mixtures. European Journal
doi.org/10.1016/j.foodres.2018.10.046. of Lipid Science and Technology, 105, 502–507.
Hatzakis, E., Dagounakis, G., Agiomyrgianaki, A., & Dais, P. (2010). A facile NMR method
Acknowledgments for the quantification of total, free and esterified sterols in virgin olive oil. Food
Chemistry, 122, 346–352.
Hatzakis, E., Koidis, A., Boskou, D., & Dais, P. (2008). Determination of phospholipids in
This research was supported by the Department of Food Science and olive oil by 31P NMR spectroscopy. Journal of Agricultural and Food Chemistry, 56,
Technology at the Ohio State University, the Food for Health Discovery 6232–6240.
Igarashi, T., Aursand, M., Hirata, Y., Gribbestad, I. S., Wada, S., & Novaka, M. (2000).
Theme and by the USDA National Institute of Food and Agriculture,

691
K. Williamson, E. Hatzakis Food Research International 119 (2019) 683–692

Nondestructive quantitative determination of docosahexaenoic acid and n−3 fatty virgin olive oil components without any sample modification. 1H NMR multi-
acids in fish oils by high-resolution 1H nuclear magnetic resonance spectroscopy. supression experiment: A powerful tool. Food Chemistry, 228, 301–314.
Journal of the American Oil Chemists' Society, 77, 737–748. Sacchi, R., Medina, I., Aubourg, S. P., Addeo, F., & Paolillo, L. (1993). Proton nuclear
de Jong, A., Plat, J., & Mensink, R. P. (2003). Metabolic effects of plant sterols and stanols magnetic resonance rapid and structure-specific determination of polyunsaturated
(Review). Journal of Nutritional Biochemistry, 14, 362–369. fatty acids in fish lipids. Journal of the American Oil Chemists' Society, 70, 225–228.
Kaffarnik, S., Ehlers, I., Gröbner, G., Schleucher, J., & Vetter, W. (2013). Two-dimensional Scharnhop, H., & Winterhalter, P. (2009). Isolation of coffee diterpenes by means of high-
31
P, 1H NMR spectroscopic profiling of phospholipids in cheese and fish. Journal of speed countercurrent chromatography. Journal of Food Composition and Analysis, 22,
Agricultural and Food Chemistry, 61, 7061–7069. 233–237.
Kamm, W., Dionisi, F., Fay, L., Hischenhuber, C., Schmarr, H., & Engel, K. (2002). Rapid Schievano, E., Finotello, C., De Angelis, E., Mammi, S., & Navarini, L. (2014). Rapid
and simultaneous analysis of 16-O-methylcafestoland sterols as markers for assess- authentication of coffee blends and quantification of 16-O-methylcafestol in roasted
ment of green coffee bean authenticity by on-line LC–GC. Journal of the American Oil coffee beans by nuclear magnetic resonance. Journal of Agricultural and Food
Chemists' Society, 79, 1109–1113. Chemistry, 62, 12309–12314.
Knothe, G., & Kenar, J. A. (2004). Determination of the fatty acid profile by 1H NMR Skiera, C., Panagiotis, S., Kuballa, T., Holzgrabe, U., & Diehl, B. (2012). 1H-NMR spec-
spectroscopy. European Journal of Lipid Science and Technology, 106, 88–96. troscopy as a new tool in the assessment of the oxidative state in edible oils. Journal of
Kovalcik, A., Obruca, S., & Marova, I. (2018). Valorization of spent coffee grounds: A the American Oil Chemists' Society, 89, 1383–1391.
review. Food and Bioproducts Processing, 110, 104–119. Somnuk, K., Eawlex, P., & Prateepchaikul, G. (2017). Optimization of coffee oil extraction
Martín, M. J., Pablos, F., González, A. G., Valdenebro, M. S., & Leon-Camacho, M. (2001). from spent coffee grounds using four solvents and prototype-scale extraction using
Fatty acid profiles as discriminant parameters for coffee varieties differentiation. circulation process. Agriculture and Natural Resources, 51, 181–189.
Talanta, 54, 291–297. Speer, K., & Kölling-Speer, I. (2006). The lipid fraction of the coffee bean. Brazilian
Merkx, D. W. H., Hong, G. T. S., Ermacore, A., Duynhoven, v., & M., J. P. (2018). Rapid Journal of Plant Physiology, 18, 201–216.
quantitative profiling of lipid oxidation products in a food emulsion by 1H NMR. Valdenebro, M. S., León-Camacho, M., Pablos, F., Gonzalez, A. G., & Martin, M. J. (1999).
Analytical Chemistry, 90, 4863–4870. Determination of the arabica/robusta composition of roasted coffee according to
Milchert, E., Smagowitcz, A., & Lewandowski, G. (2010). Optimization of the epoxidation their sterolic content. Analyst, 124, 999–1002.
of rapeseed oil with peracetic acid. Organic Process Research & Development, 14, Velazquez, P. M. C., Dieamant, G. C, Eberlin, S., Nogueira, C., Colombi, D., Di Stasi, L. C.,
1094–1101. & Queiroz, M. L. S. (2009). Effect of green Coffea arabica L. seed oil on extracellular
Monakhova, Y. B., Ruge, W., Kuballa, T., Winkelmann, O., Diehl, B., & Lachenmeier, D. matrix components and water-channel expression in in vitro and ex vivo human skin
W. (2015). Rapid approach to identify the presence of Arabica and Robusta species in models. Journal of Cosmetic Dermatology, 8, 56–62.
coffee using 1H NMR spectroscopy. Food Chemistry, 182, 178–184. Wei, F., Furihata, K., Koda, M., Hu, F., Kato, R., Miyakawa, T., & Tanokura, M. (2012). 13C
Montero De Espinosa, L., & Meier, M. A. R. (2011). Plant oils: The perfect renewable NMR-based metabolomics for the classification of green coffee beans according to
resource for polymer science? European Polymer Journal, 47, 837–852. variety and origin. Journal of Agricultural and Food Chemistry, 60, 10118–10125.
Oliveira, L. S., Franca, A. S., Camargos, R. R. S., & Ferraz, V. P. (2008). Coffee oil as a Williamson, K., & Hatzakis, E. (2017). NMR spectroscopy as a robust tool for the rapid
potential feedstock for biodiesel production. Bioresource Technology, 99, 3244–3250. evaluation of the lipid profile of fish oil supplements. Journal of Visualized Experiments
Park, J., Kim, B., & Lee, J. W. (2016). In-situ transesterification of wet spent coffee (123), e55547.
grounds for sustainable biodiesel production. Bioresource Technology, 221, 55–60. Wilson, W. K., Sumpter, R. M., Warren, J. J., Rogers, P. S., Ruan, B., & Schroepfer, G.
Plaza, D. D., Strobel, V., Heer, P. K. K., Sellars, A. B., Hoong, S., Clark, A. J., & Lapkin, A. (1996). Analysis of unsaturated C27 sterols by nuclear magnetic resonance spectro-
A. (2017). Direct valorisation of waste cocoa butter triglycerides via catalytic epox- scopy. Journal of Lipid Research, 37, 1529–1555.
idation, ring-opening and polymerisation. Journal of Chemical Technology & Xia, W., Budge, S. M., & Lumsden, M. D. (2016). 1H-NMR characterization of epoxides
Biotechnology, 92, 2254–2266. derived from polyunsaturated fatty acids. Journal of the American Oil Chemists' Society,
Raba, D. N., Poiana, M., Borozan, A. B., Stef, M., Radu, F., & Popa, M. (2015). 93, 467–478.
Investigation on crude and high-temperature heated coffee oil by ATR-FTIR spec- Zamora, R., Alba, V., & Hidalgo, F. J. (2001). Use of high-resolution 13C nuclear magnetic
troscopy along with antioxidant and antimicrobial properties. PLoS One, 10, resonance spectroscopy for the screening of virgin olive oils. Journal of the American
e0138080. Oil Chemists' Society, 78, 89–94.
Rendón, M. Y., dos Santos Scholz, M. B., & Bragagnolo, N. (2017). Is cafestol retained on Zhou, L., Khalil, A., Bindler, F., Zhao, M., Marcic, C., & Marchioni, E. (2013). Effect of
the paper filter in the preparation of filter coffee? Food Research International, 100, heat treatment on the content of individual phospholipids in coffee beans. Food
798–803. Chemistry, 141, 3846–3850.
Ruiz-Aracama, A., Goicoechea, E., & Guillén, M. D. (2017). Direct study of minor extra-

692