Headspace Analysis of the Coffee Beverage with and without Different

Chapter 26

Headspace Analysis of the Coffee Beverage with and without Different Milk Additives H . Steinhart and M. Bucking

Downloaded by PURDUE UNIV on March 25, 2016 | http://pubs.acs.org Publication Date: January 15, 2000 | doi: 10.1021/bk-2000-0754.ch026

Department of Food Chemistry, university of Hamburg, D-20146 Hamburg, Germany

M i l k and vegetable products as an additive for coffee beverages have an effect on the release of aroma substances i n the brew through their lipid, protein and carbohydrate components. For the investigation of these effects an external dynamic headspace sampling technique and an external static headspace technique were developed. The release of volatile compounds of the beverages i n the oral cavity of human volunteers was measured by oral vapour gas chromatography. With these techniques the most potent odorants of the coffee beverage were determined. Analyses were performed by gas chromatography/olfactometry, FI- and M S – detection. To characterize the odour profiles of the different beverages G C / O analysis were used. All beverages with a milk or vegetable additive showed reduced, but typical odour profiles for each additive.

Introduction U p to now, investigations of the coffee flavour have been confined to the analysis of the aroma substances (1, 2, 3). These investigations showed that about 30 volatile compounds were substantially responsible for the coffee flavour. The aim of this study was to investigate the influence of different milk additives and one coffee whitener on the release of flavour impact compounds from coffee beverages. K I M et al. (4) were the first to investigate the effects of milk additives on the coffee flavour; they used the conventional static headspace technique and only instant coffees. The purposes of adding these products to the coffee beverage are; to develop a desirable colour change, to impart a body to the coffee beverage, to reduce bitter and sour tastes and to reduce astringency of the coffee.

252

© 2000 American Chemical Society

Parliment et al.; Caffeinated Beverages ACS Symposium Series; American Chemical Society: Washington, DC, 2000.

253 Ingredients of these additives such as lipids, proteins and carbohydrates affect the retention of volatiles (5). Consequently, these aroma interactions affect quality and quantity of the coffee headspace aroma. Direct injection of a headspace sample onto a G C column gives the most accurate composition of flavours. However, working with static headspace (gastight syringe) was not sufficient, because only low amounts of aroma compounds were collected. Therefore new devices had to be developed, which could collect a larger volume of headspace above the coffee beverage. Static headspace as distinct from dynamic headspace measures the concentration of volatiles under equilibrium conditions. In the present study both methods were carried out by G C - F I D / olfactometry and G C - M S / olfactometry. Furthermore oral vapour gas chromatography ( O V G C ) was performed to charaterise the human aroma impression.


Materials and Methods Sample Material For the beverages the two economically important coffee species were used: one Arabica coffee (Columbia) and one Robusta coffee (Indonesia), both with an average roasting degree. Besides this a soluble coffee was used. A l l samples were supplied by Kraft Jacobs Suchard (Bremen, FRG). Eight products, purchased from a local market, respectively Kraft Jacobs Suchard (Munich, FRG), were selected as typical coffee additives. These different types of dairy and vegetable products were also chosen because of their different lipid and protein contents. These components have the greatest influence on the retardation of volatiles. The ingredients of the additives are listed in Table I (6).

Table I. Ingredients of the additives (in %) Additive UHT-Milk Condensed milk Coffee creamer Whipping cream Coffee whitener (vegetable)

Lipid 3.5 10 10 30 34

Carbohydrate 4.8 12.5 3.1 3.2 55

Protein 3.3 8.8 4.0 2.5 6

Sample Preparations External Dynamic Headspace Sampling (DHS) The beans were stored at -17°C and ground directly prior to use in a coffee grinder of the style that is normally used in coffee shops. The brew was prepared in a household coffeemaker with 12g coffee powder and 225g tap water. Soluble coffee was prepared by pouring boiling water (125g) onto the powder (3.5g). 50g and 125g of soluble coffee brew respectively were placed in the external


254 dynamic sampling device (Figure la), with a headspace volume of 275mL, and an additive was added (10g and 25g respectively). The temperature of the water bath was 40°C. With a flow of 40mL/min for 30 minutes, nitrogen was flushed above the coffee and the volatiles were collected on Tenax T A tubes.


External Static Headspace Sampling (SHS) For the external static headspace device 220g of coffee beverage and 45g of additive were used, These liquids were filled in the lower of the two glass vessels (Figure lb). After an equilibrium time of 15 min at room temperature the lower vessel was replaced by an empty one. Nitrogen was flushed with lOOmL/min for 30 minutes and the volatiles were collected on a Tenax T A tube. Standard deviation of 8 replications ranged between 2 and 10% for most of the volatile compounds for both methods.

lb

la

Tenax Tube Tenax Tube

Nitrogen

1

Nitrogen

Drying agent •1

Drying agent

Volume: 850ml

Water bath

- * Coffee Coffee

Figure la and b. External Dynamic Headspace device (la) External Static Headspace device (lb)

Oral Breath Sampler For the O V G C five assessors took 18mL of the freshly brewed beverage i n their mouth, the beverage remained i n their oral cavity. Their breath was collected using the Oral Breath Sampler (7) as shown i n Figure 2. For this they had to place a mouthpiece between their lips i n such a way that released volatile compounds were directed onto a Tenax T A tube by a vacuum pump with a constant flow (142mL/min) for 6 minutes. A cold trap (-10°C) was used for freezing out water vapour.


255 Reproducibility of this method was checked by analysing the same kind of beverage five times with subsequent comparison of the chromatograms.

mouthpiece


vacuum pump

Figure 2 Oral Breath Sampler (J. ROOZENandA. LEGGER-HUYSMAN, (Reproduced with permission from reference 7. Copyright 1994.)

1994)

GC-FID, GC-MS and GC-Olfactometry DHS/SHS The volatile compounds were analysed on an H P 5890 G C equipped with an FI-Detector and Sniffing-Port or MS-Detcctor and Sniffing-Port. The volatiles were desorbed by a thermal desorption device and injected onto a DB-5 capillary column (30m χ 0.53mm; 1.5um film thickness, non-polar). In addition to the DB-5, an OV1701 (60m χ 0,32mm; 1.0pm film thickness or 60 χ 0.25mm; 0.5μηι film thickness, semi-polar) and an F F A P (60m χ 0.25mm; 0.5μηι film thickness, polar) fused silica capillary column were used. Oral breath sampler The volatile compounds were analysed on a Carlo Erba M E G A 5300 G C equipped with an FI-Detector and/or a sniffing-port. The volatiles were desorbed by a thermal desorption device and injected onto a Supelcowax 10 capillary column (60m χ 0.25mm; 0.25μηι film thickness). Identification of volatiles The identification of the compounds was achieved by comparison of retention data on D B - 5 , OV-1701 and F F A P and mass spectral data as well as sensory properties with those of authentic reference substances. Mass spectra were generated at 70eV i n the electron impact mode.


256 G C / O analysis


The assessors recorded the aroma substances during the sniff runs (Figure 3): When the compound exceeds it's threshold, the sniffer records (chart speed: lOcm/min) the event; when the concentration drops below the threshold again, the sniffer also records the event. Flavour descriptors were generated during preliminary GC / sniffing experiments and clustered to descriptors after group sessions of the panel (five trained assessors).

Spider-web diagram Figure 3. Creation of odour profiles with a GC/O-technique

Results

Comparison between DHS and SHS and the influence of additives on the intensity of coffee volatiles using DHS and SHS analysis With both methods over 50 potent odorants (Table 2), resulting in a list of ten descriptors, were recognized at the sniffing port. Collecting volatiles with the DHS affected slightly higher amounts of these volatiles. The identification of these volatiles verified most of the contributors of the coffee aroma described in literature (1, 2, 3). The results of the GC/O-analysis showed that both odour profiles were comparable. Thus the influence of non-equilibrium conditions with the DHS was negligible. In general the additives reduced the intensity of the volatiles (Figure 4), especially for the roasty descriptor. Also the descriptors malty, cocoa, and fruity, flowery showed a significant decrease. The changes in the flavour profile of coffee beverages with an additive can be caused by several effects (8). Interactions (e.g. solution, adsorption, specific binding) of the volatiles with the ingredients of these


257

Table 2. Examples of identified potent odorants


Compound

Retention index on Aroma quality OV1701 DB-5 putrid

Headspace Analysis of the Coffee Beverage with and without Different

Recommend Documents