Caballero B. (ed.) Encyclopaedia of Food Science, Food Technology and Nutrition. Ten-Volume Set

Подождите немного. Документ загружается.

Extraction

0013 Percolation batteries with five to eight interconnected

columns holding roast coffee in different stages of

exhaustion is the most widely used system for extrac-

tion to obtain coffee extracts of an economic soluble

solids concentration (20–25% w/w) for subsequent

drying. Its operation and variants are fully described

in the literature, though it should be emphasized that

the operation is intermittent (for draw-off of extract),

countercurrent (in respect of water–coffee), and under

pressure for all columns except the first (to keep the

system hydraulic at temperatures in excess of 100



up to 175



C). The first column contains fresh roasted

coffee extracted with liquor at 100



C from which

draw-offs are made, whilst the most spent is contacted

with pure feed water at the highest temperature. The

operation is outlined in Figure 2. Recent patents sug-

gest the use of ever higher temperatures for very short

times in the final stages of extraction.

Concentration of Extract

0014 The removal of water from a percolated coffee ex-

tract is more economically achieved if concentrated

extract is fed to the driers. Evaporation is widely used

and again, for reasons of heat economy, conducted in

various types of multistage units. Short-contact-time

evaporators are favored, such as plate and centrifugal

film evaporators, coupled with close attention to op-

erating temperatures to minimize undesirable changes

in the extract. However, evaporation of water is

always accompanied by evaporation and loss of

organic volatile substances contributing to flavor.

Certain prestripping techniques are available to over-

come this problem.

0015 Freeze concentration can be used alternatively,

which has the marked advantage of substantially

retaining the volatile substances whilst the water is

removed as ice, but has the disadvantage of only

allowing a final concentration of about 38% w/w

soluble coffee solids, due to high viscosity problems.

Spray Drying

0016 Extracts from a percolation battery as described

above can be directly spray dried, though an inter-

mediate filtration stage (centrifuges) may be included.

To provide dried particles of an average size around

300 mm, needed for a satisfactory bulk density and

flowability, specialized spray driers are used. The

main features are:

0017 tall drying chambers, e.g., of height 7.5 m or more;

0018 use of centrifugal pressure nozzles rather than

spinning disks for feed spraying, with their

inherent narrow angle of discharge;

0019internal chamberseparationarrangements between

desired product and dust, though not essential.

Spray drying is conducted also to give powder of

desired colour, and a final moisture content of less

than 5% w/w.

Time zero

Coffee in

Feed water in

Feed to fresh on 6

Fill column 1

Vent at column 6

Start draw on 6

Isolate column 2

Blow down column 2

Finish draw on 6

Feed to fresh on 1

Time one cycle

Feed water in

2 3 4 5

fig0002Figure 2 Typical sequence of events in the operation of a

percolation battery for instant coffee. The shape, size, and

number of columns are diagrammatic. Reproduced from Clarke

RJ and Macrae R (1987), with permission from Kluwer Academic.

1496 COFFEE/Instant

0020 More concentrated extracts, say up to 60% w/w

soluble solids, may also be spray dried, though some

controlled foaming of extract with carbon dioxide or

nitrogen gas is then generally required, if the desired

physical properties are to be maintained. (See Drying:

Spray Drying.)

Freeze Drying

0021 Freeze drying has been particularly successful for

coffee extracts, when its use for other foodstuff

liquids/solids has markedly declined in the last

decades. However, a specialized technique has been

developed, first patented around 1965, whereby the

coffee extract is first frozen, and then the slabs are

granulated, whilst still frozen, to particles approxi-

mately the same size as desired in the finished dried

product. Oversize/undersize particles are recycled.

There are a number of designs of freeze drier avail-

able, which will generally handle the frozen granules

in trays resting on heated shelves in a batchwise

manner. The amount of time of freeze drying required

is up to 7 h under a very high vacuum (approx 0.4

torr), and a carefully controlled supply of heat to the

drying granules, by conduction and/or radiation.

0022 Whilst satisfactory product can be obtained by

freeze-drying extracts direct from a percolation

battery, it is more usual for economic and other

reasons for concentrated extracts up to 40% w/w by

freeze concentration (q.v.) or by evaporation (40%

w/w or higher) to be freeze dried. If a favorable bulk

density of 180–220 g l

1

is required, it is additionally

necessary to foam the extract in the slush-frozen

form, before full freezing and freeze drying. (See

Freeze-drying: The Basic Process.)

Separate Volatile Compound Handling

0023 In recent decades, instant coffee manufacture has

become increasingly sophisticated, through greater

attention to methods of maximizing extraction and

retention of volatile compounds responsible for

flavor/aroma, which earlier procedures of simple

extraction and spray drying could not accomplish.

The first method concerns headspace aroma of the

finished powder, due to the presence of very highly

volatile substances that cannot be retained by any

method of spray or freeze drying. A convenient

vehicle was found to be coffee oil, which can be

sprayed or plated (at a level of about 0.5% or less)

on to the powder/granules as already described, and

will give an aroma similar to that of dry roasted and

ground coffee when sniffed. The coffee oil can be that

obtained by mechanical expression of part of the

roast coffee blend to be percolated, or a spent grounds

coffee oil purified and enhanced with suitable

aromatics from other stages of the manufacturing

process. The coffee oil and its aromatics are very

susceptible to oxidation, hence the need for gas-

packed product, as already described. Its application

(so-called ‘aromatization’) will play only a small role

in the flavor of the made-up beverage product, so that

other methods are needed, of which there are many

available, mostly patented. (See Oxidation of Food

Components.)

0024Battery extraction, designed primarily for the ex-

traction of soluble solids, can be fairly successful also

for the extraction of volatile compounds, so that,

when followed by freeze concentration and freeze

drying, a flavorful product is formed. However, vola-

tile compounds can be prestripped by steam from the

roasted coffee before aqueous extraction, in the form

of an aqueous condensate to be reincorporated later.

Similarly, a percolated extract can be stripped of the

important volatile compounds by partial evaporation

(say 10–20% of the water), again to give an aqueous

essence condensate. The stripped extract can then be

further evaporated as required, with the condensate

discarded.

0025The final solution after evaporation containing a

high concentration of solubles, can then be slightly

diluted back with either or both of the aqueous

essence condensates, and used as feed for either

spray or freeze drying. The retention of volatile sub-

stances in spray drying is known to be very markedly

improved when feed extracts of high soluble solids

concentration are spray dried. This will be true also of

freeze drying, though, additionally, it has been found

that parameters of the freezing itself are also very

important (e.g., slow freezing to large ice crystals

favors subsequent volatile retention).

Agglomeration

0026Agglomeration was a process first used for dusty skim

milk powders to increase average particle size and,

more importantly, for ease of rapid dissolution in

water, so-called ‘instantization.’ Whilst the latter is

not required for a satisfactory instant coffee powder,

a similar agglomeration process, using steam/water to

rewet the surface of particles followed by drying, is

used to manufacture instant coffee granules that are

not in fact freeze-dried. Precise details of processes

differ, as the numerous patents will indicate. The

diminution of volatile compound content from spray

drying to granule formation will only be small.

Packing

0027Instant coffee was originally packed into tins but,

since 1960, has now been almost entirely packed

into glass jars. Under EEC prescribed weight

COFFEE/Instant 1497

directives, allowable packed weights for retail sale are

50, 100, and 200 g, and certain multiples, accompan-

ied by the average weight system to be assessed over a

stipulated number of containers. Filling machinery is

now very high-speed (e.g., 250 per minute for 50 g

jars), operating by volumetric fill, and adjustable by

vacuum level according to the bulk density of the

product being packed. Two other operations may

also be incorporated, simultaneous spraying of small

amounts of coffee oil into the powder/granules for

headspace aroma purposes and a final provision of a

carbon dioxide/nitrogen atmosphere within the jar to

lower the headspace oxygen level to 4% v/v or

preferably less.

General Comments

0028 In recent decades, all these process stages have

been investigated fundamentally through the discip-

lines of chemical (food) engineering by a number of

people, especially the late Professor Thijssen of Eind-

hoven Technical University, to provide considerable

mathematical insight into their mechanisms.

Storage and Stability

0029 Instant coffee is relatively hygroscopic, easily picking

up moisture from the atmosphere and caking at about

7–8% w/w moisture content. For example, to keep

instant coffee below 5% (dry basis) moisture content,

the relative humidity of the air with which it is in

contact must be below 35–40%, though the precise

value depends upon the nature of the instant coffee in

question, primarily due to differences in porosity of

the particular particles made. It is necessary, therefore,

that jars of instant coffee be well sealed prior to sale.

0030 A simple spray-dried product, provided the dry

matter content is kept above 95% w/w (i.e., < 5%

w/w moisture content), will have a shelf-life of several

years. The modern sophisticated products, containing

substantial amounts of volatile compounds (espe-

cially plated coffee oils), need to be packed in jars

also with no more than 4% in-pack oxygen content

and preferably less, so that the shelf-life to acceptable

quality is maintained up to 18 months.

See also: Caffeine; Carbohydrates: Classification and

Properties; Drying: Spray Drying; Evaporation: Basic

Principles; Uses in the Food Industry; Freeze-drying: The

Basic Process; Structural and Flavor (Flavour) Changes;

Oxidation of Food Components

Further Reading

Clarke RJ (1987) Coffee technology. In: Herschdoefer S

(ed.) Quality Control in the Food Industry, 2nd edn,

vol. 4. London: Academic Press.

Clarke RJ and Macrae R (1985–1988) Coffee: Chemistry,

vol. 1, Technology, vol. 2, Commercial and Technico-

legal Aspects, vol. 6. Barking, UK: Elsevier.

Clarke RJ and Vitzthum OG (2001) Coffee: Recent Devel-

opments. Oxford: Blackwell Science.

Clifford MN and Willson KC (eds) (1985) Coffee: Botany,

Biochemistry, Production of Beans and Beverage.

London: Chapman & Hall.

Analysis of Coffee Products

L C Trugo, Universidade Federal do Rio de Janeiro,

Rio de Janeiro, RJ, Brazil

Introduction

0001The composition of the coffee seed is complex, as is

the case for other natural products. After roasting at

unusually high temperatures many chemical reactions

take place, with degradation of the original com-

pounds as well as polymerization of intermediate

products, which substantially increases its complex-

ity. These reactions are indispensable for the charac-

teristic coffee beverage. Analysis of the different

coffee products may be relatively simple, involving

inexpensive equipment such as that used for proxim-

ate analysis, or may demand a great deal of analytical

expertise and sophisticated apparatus, such as chro-

matography coupled to mass spectrometers, infrared

spectrometers, and capillary electrophoresis to assess

coffee flavor composition and authenticity.

Proximate Analysis

0002Proximate analysis in food products is the analysis of

the major constituents which together comprise

nearly 100% of the food composition. This generally

includes water, ash, total protein, and carbohydrates.

Proximate analysis is the first approach for food

product characterization and control and it is not

dependent on expensive or sophisticated equipment.

It is easily carried out by quality control laboratories,

including coffee-processing plants.

Water

0003The control of the water content of coffee at different

stages of processing, from green beans to instant

coffee, is an important action to achieve a desirable

final product. In the first stages of processing the ripe

coffee cherries, containing 60–65% water, must be

dried to a level of 10–12% which is normally found

in the green beans, and to prevent undesirable

1498 COFFEE/Analysis of Coffee Products

microbial growth which can alter the sensory quality

of the final product. After roasting, the water level

drops to about 5% and after industrial extraction and

drying to produce instant coffee, the powder should

present values in the range 2–4% of water. The sim-

plest method for general moisture determination in

food is oven drying at 105–110



C until constant

weight is reached, and this may be applied to coffee

products. However, some inaccurate data may be

obtained, particularly when roast or instant coffee is

used, due to water formation by the browning

reaction or by loss of volatile components at these

conditions.

0004 Empirical time–temperature procedures have been

proposed, including a two stage drying process for

green coffee samples, involving a drying period at

130



C for 6 h, followed by another of 4 h after a

rest period of 15 h in a desiccator. This procedure is

recognized as giving reproducible data; however, it

may not be convenient for routine analysis. Vacuum

oven-drying methods are also recommended, particu-

larly for roast and instant coffee, because it is possible

to use lower temperatures of around 60



C. However,

in practice the overnight oven-drying method at

105



C is still adopted. Nondestructive methods

based on nuclear magnetic resonance, near infrared

spectroscopy, and microwave techniques may also be

used to assess the water content of coffee samples but

their wide application may be limited due to the

relative high cost. The alternative for more accurate

water determination is the use of the Karl–Fischer

method based on a specific quantitative chemical

reaction of the water from the sample. Automated

equipment based on the Karl–Fischer reaction is

commercially available and may be useful for routine

purposes.

Ash

0005 Ash determination can easily be achieved by dry ash-

ing the coffee sample at 550–580



C for at least 6 h

and usually up to 17 h to obtain constant ash weight.

The ash content of green coffee is in the range 3.0–

4.5%. This range does not alter significantly in

roasted coffee but increases considerably in instant

coffee, with values in the range 9–10%. The ash

content gives a rough idea of the total mineral

amount and it has been a useful parameter for coffee

standardization and to detect instant coffee adulter-

ation by the addition of coffee substitutes.

Crude Protein

0006 The crude protein content of coffee products is usu-

ally assessed by the universally used Kjeldahl method

for total nitrogen determination. Together with other

data of proximate analysis this may be useful for

quality control but is of very little use as an indicator

of a food source of protein since coffee products are

not relevant as protein sources in the diet; however,

they are important for coffee aroma formation during

roasting. The total protein content of green coffee is

in the range 11–13% and this value is increased after

roasting to 13–15% and 16–21% in the instant

coffees. Although these data may be useful for quality

control, one must consider that total nitrogen

includes non-nitrogen proteins, particularly from

trigonelline and caffeine which will inflate the actual

protein results. Instant coffees analyzed only for

protein nitrogen can present final protein values in

the range 8–13% and if the actual protein value

is needed, the non-nitrogen content should be

subtracted from the crude protein content.

Oil

0007Total lipids as determined by organic solvent extrac-

tion are present in high levels in green and roasted

coffees. Significant differences are found in different

species, with Arabica coffee containing about 12–

13% and Robusta 7–10%. They are important

vehicles for coffee aroma and are relatively stable

during roasting, showing a relative increase in roasted

coffees: Arabica presents 14–20% and Robusta 11–

16%. However, only minute amounts are found in the

beverage or even in instant coffee since they are not

very soluble during domestic or industrial hot-water

extraction. The saponifiable fraction of the green

coffee oil is the major component, with triglycerides

accounting for 70–80% and free fatty acids 0.8–2%.

The traditional Soxhlet procedure, based on organic

solvent extraction, for example with petroleum ether,

is usually applied for routine analysis; however, for

green coffee samples it may be necessary to carry out

a previous acidic hydrolysis for liberation of structur-

ally bound lipids in order to obtain more accurate

results.

Total Carbohydrates

0008The carbohydrate composition of coffee products

varies considerably during processing. In the green

coffee bean large amounts of both sugars and poly-

saccharides are found. In roasted coffee only polysac-

charides are found since sucrose is destroyed by

caramelization during roasting. However, in instant

coffees, free sugars are again present, and these

are mainly monosaccharides, due to hydrolysis of

polysaccharides during industrial extraction. Colori-

metric methods such as the phenol-sulfuric or the

anthrone procedures may be applied to assess the

total amount of carbohydrates, in green, roasted, or

instant coffees; however, a great degree of interfer-

ence is expected, particularly in roasted and instant

COFFEE/Analysis of Coffee Products 1499

coffee, due to complex browning components formed

during processing. Due to this general difficulty many

laboratories simply calculate the total carbohydrates

by substracting the total proximate results from 100%.

Protein and Amino Acids

0009 Coffee presents a relatively high level of protein in the

range 11–16% in the green bean with less than 0.5%

corresponding to free amino acids. However, coffee is

not considered to be a source of protein in the diet

since most of this material will remain in the insoluble

residue after water extraction to prepare the coffee

beverage. The nitrogen-protein determination will

give only a crude idea of the protein content. More

accurate results may be obtained by amino acid

analysis. The study of specific protein fractions may

be achieved by gel permeation or ion exchange chro-

matography, hydrophobic and reverse-phase chroma-

tography, or by electrophoresis. The importance of

protein and free amino acids is related to their contri-

bution to coffee aroma compounds formation during

roasting and also to pigment formation though the

Maillard reaction. The degradation of amino acids

during roasting may lead to the formation of import-

ant volatile compounds such as pyrazines, oxazoles,

and pyrroles. The analysis of amino acids by acid

hydrolysis of protein will provide more accurate

true protein data in green coffee. However, for

roasted and instant coffees acid hydrolysis will liber-

ate amino acids from Maillard condensed material

with little relation to true protein composition. For

tryptophan quantification a specific colorimetric

method after basic or enzymatic hydrolysis should

be applied, since this amino acid is not resistant to

the current acid hydrolysis. One less common amino

acid, pipecolic acid, has been used for coffee species

characterization since it is present in Coffea arabica

but not in C. robusta.

0010 Amino acid analysis may be carried out using

dedicated amino acid analyzers based on ion ex-

change chromatography followed by postderivatiza-

tion with ninhydrin and detection by colorimetry.

More recently, high-performance liquid chromatog-

raphy (HPLC) using reverse-phase columns with

prederivatization has been used. The protein hydro-

lysate may be treated with phenyl isothiocyanate to

form phenylthiocarbamyl derivatives followed by

chromatography and ultraviolet detection or submit-

ted to reaction with o-phthaldialdehyde (OPA) to

produce OPA derivatives for fluorimetric detection.

Capillary electrophoresis has also been applied and

shows great potential as a competitive technique in

relation to sensitivity and resolution. Free amino acid

analysis may be important as aroma precursors and

their determination can be achieved by deproteina-

tion with 10% trichloroacetic acid and the amino

acid analysis of the supernatant, as described above.

(See Amino acids: Properties and Occurrence.)

Free Sugars and Polysaccharides

0011More sophisticated techniques are needed for a more

detailed carbohydrate analysis. The determination of

individual free sugars has become very important,

particularly in instant coffee samples, since the sugar

profile seems to be an accurate indicator of coffee

authentication. Sucrose is the main sugar present in

green coffee and it may easily be determined by HPLC

with normal-phase chromatography (generally silica-

amino), acetonitrile (70–80%) in water as mobile

phase and refractive index detection, after hot water

or aqueous ethanol (80%) extraction, in the ground

and defatted beans. The extract should be cleaned up

by means of Carrez solutions or activated charcoal

and filtered before chromatography. The extract may

be concentrated if necessary, particularly when

roasted and instant coffees are used. However, due

to rapid degradation of sucrose during roasting and

the appearance of smaller amounts of other free

sugars due to polysaccharide hydrolysis during

instant coffee manufacturing, the analysis of sugars

in both roasted and instant coffees is much more

complicated. Chromatographic analysis may be diffi-

cult with the use of refractive index detector due to its

low sensitivity. The application of evaporative mass

detectors and pulsed amperometric detectors has

produced better results.

0012The polysaccharide fraction of coffee is important

as a substrate for pigment and aroma formation and

also as a source of free sugars, particularly in instant

coffee production. The industrial extraction at high

temperatures during instant coffee manufacture

promotes some degree of polysaccharide hydrolysis,

liberating constituent monosaccharides and oligosac-

charides from the polysaccharide backbone. Conse-

quently, arabinose, galactose, mannose, and glucose

from arabinogalactan, galactomannan, and cellulose

are typical of instant coffees. The analysis of poly-

saccharides is mainly studied after separation of

fractions by solvent extraction or by gel permeation

chromatography and the subsequent analysis of the

constituent monosaccharides after acid hydrolysis

(1 mol l

1

sulfuric acid) by gas chromatography or

HPLC. Table 1 shows the sugar composition of the

water-insoluble fraction from green coffee which con-

tains most of the polysaccharides, before and after

being submitted to roasting conditions. Arabinose

shows greater lability to heating and this may be

explained by the presence of short side chains with

1500 COFFEE/Analysis of Coffee Products

terminal arabinose in the arabinogalactan structure.

Due to this characteristic, arabinose from the poly-

saccharide fraction decreases significantly during

roasting and is found in relatively high levels in

instant coffee.

Lipids

0013 Lipids are present in green coffees at relatively high

levels (7–13%); higher levels are usually found in

Arabica coffee. They are relatively stable during

roasting, showing a relative increase in roasted coffee

(11–20%). The saponifiable fraction is predominant,

with triacylglycerols accounting for over 70% of the

coffee oil. The analysis of constituent fatty acids is

achieved by usual transmethylation as a derivatization

procedure of triacylglycerols, prior to gas chromatog-

raphy of the fatty acid methyl esters. The fatty acid

composition of coffee oil shows linoleic acid as the

main component (45% of total fatty acids) followed

by palmitic (35%), oleic (8%), stearic (7%), arachidic

(3%), and linolenic (2%). The nontriacylglycerol frac-

tion is quite different from other oil seeds. High levels

of diterpenes and diterpene mono-fatty acid esters are

found, particularly cafestol and kahweol, which ac-

count for about 20% of the oil fraction. Apparently,

kahweol is present in Arabica but absent in Robusta

coffee and for this reason it has been used for identifi-

cation of coffee species. Cafestol and kahweol have

received much attention recently since it has been con-

firmed that they can increase plasma cholesterol levels,

as observed experimentally and by epidemiological

studies. Curiously, the correlation between coffee con-

sumption and high levels of cholesterol has been found

only in the Scandinavian population drinking unfil-

tered coffee. It was observed that unfiltered coffee

has small amounts of a floating oil material on the

surface. This material was determined to be mainly

cafestol and kahweol, which were proved to be chol-

esterol-raising factors in experimental animals. Coffee

filtration appears to be sufficient to retain all the diter-

penes in the filter, explaining why the correlation

between high cholesterol and coffee drinking is not

verified in consumers of filtered coffee. The isolation

of diterpene alcohols and their esters has been carried

out by paper, thin-layer, or column chromatography.

The qualitative detection of kahweol is based on the

reaction with potassium iodide (KI) in acetic acid

medium; this produces a blue color after heating.

Sterols are also important components of the unsapo-

nifiable fraction, accounting for some 20% and

consisting mainly of sitosterol, stigmasterol, and cam-

pesterol. Terpenes and sterols may be analyzed by gas

chromatography after deesterification. To aid in struc-

ture elucidation, the use of gas chromatography–mass

spectrometry may be helpful. The analysis of coffee

wax has been carried out to detect the presence of 5-

hydroxytryptamine, which appears to be an irritant of

the gastric mucosa. The use of reverse-phase HPLC

seems to be adequate for the analysis of this class of

compound. Because coffee beverage is an aqueous

extract, only a small amount of lipid will be dispersed

in the liquid and only those with specific biological

activity will have a greater impact for the consumer.

Minerals

0014More than 30 minerals have been determined in

coffee products. An unusually high potassium content

is found in both Arabica and Robusta coffees, repre-

senting about 40% of total mineral content. Particu-

larly interesting is the fact that Arabica coffees

present consistently a higher amount of manganese

when compared to Robusta, indicating a genetic dif-

ference. In instant coffees the mineral profile is simi-

lar to green coffee; however, due to salt solubility in

water the total mineral content will be higher than

green and roasted coffees. The control of ash and

also potassium contents of instant coffees may

be useful as an indicator of extraction yield. The

analysis of potassium may be easily carried out by

inexpensive flame photometers. A more complete

assessment of mineral composition may be achieved

using more comprehensive techniques such as

atomic absorption, thermal neutron activation,

and plasma spectrometry or inductively coupled

plasma mass spectrometry (ICP-MS), after sample

mineralization.

Trigonelline and Niacin

0015Trigonelline is a pyridine derivative present in several

species of fruits and seeds, including coffee. It is pre-

sent in both Arabica and Robusta coffees in the aver-

age amount of 1%. It is extensively degraded during

coffee roasting and when dark roasting conditions are

used, only about 0.1–0.2% remains in the roasted

tbl0001 Table 1 Composition of water-insoluble polysaccharide from

Arabica coffee before and after roasting

Sugarsin the

polysaccharide fraction

Nonroasted

g% (dry basis)

roasted

g% (dry basis)

Arabinose 11.0 6.0

Galactose 24.8 23.9

Mannose 38.0 41.7

Glucose 20.0 22.0

From De Maria CAB, Trugo LC, Aquino Neto FR, Moreira RFA and Alviano

CS (1996) Composition of green coffee water-soluble fractions and

identification of volatiles formed during roasting. Food Chemistry 55:

203–207.

COFFEE/Analysis of Coffee Products 1501

coffee (Table 2). Due to the high solubility in water,

trigonelline is readily extracted from instant coffees,

appearing in the range 0.9–1.7%. The final concen-

tration in instant coffee depends on the extraction

conditions used as well as on the original material

and on the roasting degree. Many volatiles are formed

by trigonelline degradation, such as pyrrols and

pyridines which will contribute to the aroma fraction

of the coffee. Nicotinic acid is another important

product formed by demethylation of trigonelline.

Nicotinic acid or niacin is a vitamin which is absent

in green coffee and in fact is produced during coffee

roasting, being present in both roasted and instant

coffees, in the range 10–26 and 20–50 mg%, respect-

ively. Because niacin is produced during coffee pro-

cessing, the vitamin is highly bioavailable in the

coffee beverage, in contrast with natural sources

where it is present in the bound form.

0016 The analysis of trigonelline in coffee products may

be carried out by isolating it from coffee extract using

a celite column followed by measurement by ultra-

violet (UV) spectrophotometry, by paper chromatog-

raphy followed by elution and UV measurement of

the spot, or by HPLC. This last technique has been

shown to produce more reliable results and is pres-

ently widely used. A simple reverse-phase method

using a C

column, water with a small amount of

methanol as solvent, and monitoring at 265–272 nm

has been quite adequate for routine analysis. Clarifi-

cation of coffee extracts may be achieved by solid-

phase extraction or precipitation of proteins and pig-

ments by means of lead acetate solution or other

adsorbent material except Carrez solutions, since tri-

gonelline is strongly adsorbed by zinc ferrocyanide.

0017 Nicotinic acid may be determined by microbio-

logical, colorimetric (with cyanogen bromide and p-

aminoacetophenone), or chromatographic methods.

However, HPLC has frequently been the technique of

choice. In this case better results have been achieved

using reverse-phase procedures with a C

column,

and a buffered ion-paired mobile phase to attenuate

the nicotinic acid polarity and increase its retention

in the column and detection at 254 nm. As for

trigonelline analysis, some clarification procedure of

the coffee extract must be used.

Caffeine

0018Caffeine is widely consumed in many countries as a

constituent of many food items such as coffee prod-

ucts, tea, cocoa products, and cola beverages. Due to

its physiological and pharmacological effects, caf-

feine is the most studied compound present in coffee.

Caffeine is the major alkaloid present in coffee, al-

though low levels of theobromine and theophylline,

which are other purine alkaloids, may also be found.

Caffeine is the component responsible for the stimu-

lant properties of coffee beverages. Green Robusta

coffee has a higher content of caffeine (1.5–2.5%)

than Arabica (1.0–1.2%) and it is relatively stable to

heat, showing a small loss during the roasting pro-

cess. Due to its solubility in hot water, caffeine is

highly extracted during instant coffee production.

The amount present in instant coffee strongly

depends on the green coffee blend used and less so

on extraction rate. Robusta coffee is widely used as a

blend component in instant coffee manufacture and

because it has considerably more caffeine than Arab-

ica the final content depends on the proportion of

these types of coffees. Usual figures would fall in the

range 2.5–4.8% for regular instant coffee and 0.2%

for decaffeinated coffee. (See Caffeine.)

0019The analysis of caffeine was carried out for many

years using the method based on its extraction from

the coffee matrix with chloroform, followed by

extensive purification of the extract with celite or

alumina and subsequent measurement of caffeine ab-

sorption at 272 nm. However, extensive extract puri-

fication procedures are needed to avoid the presence

of significant amounts of interfering materials. The

application of chromatographic techniques, particu-

larly HPLC and gas chromatography, has produced

more reliable results. HPLC has been widely

adopted due to its simplicity, rapidity, and reliability.

tbl0002 Table 2 Loss of trigonelline during coffee roasting and the correspondent increase in niacin

Typeof coffee Type of roasting Trigonelline g% (dry basis) Niacin mg% (drybasis)

Arabica Light 0.9 4.9

Medium 0.7 10.2

Dark 0.4 14.2

Very dark 0.2 15.0

Robusta Light 0.7 9.9

Medium 0.6 15.0

Dark 0.3 23.4

Very dark 0.1 27.0

From Trugo, LC (1984) PhD thesis. University of Reading, UK.

1502 COFFEE/Analysis of Coffee Products

Reverse-phase HPLC methods using a C

column

and aqueous methanol as mobile phase with UV

monitoring at 272 nm have proved to be perfectly

adequate for fundamental and quality control ana-

lyses. The water solution obtained from the coffee

matrix may be easily purified by solid-phase extrac-

tion or by clean-up using Carrez solutions prior to

chromatography. When analysis of a large amount of

samples is continuously needed, an HPLC method

may be relatively easily automated.

Chlorogenic Acids

0020 Chlorogenic acid is a trivial name used to define the

major phenolic components found in coffee. This name

originates from early observations that green coffee

extracts produced a green color after addition of a

solution of ferric chloride. In fact, chlorogenic acids

present in coffee are mainly esters of quinic acid with

caffeic, ferulic, or coumaric acids. Significant amounts

of three isomers (in positions 3, 4, and 5 of quinic acid)

of each of these groups of caffeoylquinic, di-caffeoyl-

quinic, and feruloylquinic acids are usually found in

green and processed coffees. They are not resistant to

heat and are degraded during roasting, contributing to

the aroma fraction. Consequently, their levels in the

coffee beverage are dependent on the roasting degree,

with less than 10% of the original green coffee content

found in very dark roasted coffees (Table 3).

0021 Chlorogenic acids and their derivatives have re-

ceived much attention due to recent investigations

showing different biological properties of this class

of compounds. A lactone of ferulic acid presented

affinity to bind to opioid receptors in the brain. The

antioxidant activity of caffeoylquinic acid has been

demonstrated and di-caffeoylquinic acids presented

strong inhibitory activity against integrase enzymes,

which are important for virus replication. Although

the biological significance of the ingestion of coffee

chlorogenic acids is not well defined, their in vitro

activities observed in different studies are sufficient to

draw attention to the chemical and biological charac-

teristics of these molecules.

0022Colorimetric and chromatographic methods have

been applied to chlorogenic acid analysis. The first

approach is used for determination of the caffeoyl-

quinic group by reaction with molybdate or total

determination using the metaperiodate reagent. The

latter is useful for individual chlorogenic acid isomer

analysis. Although gas chromatography has been ap-

plied to chlorogenic acid analyses, HPLC appears to

be the preferred technique. Chlorogenic acid extrac-

tion is critical for obtaining accurate results. Different

solubilities of different isomers can make quantitative

extraction difficult. Hot solvent extraction appears to

give more reproducible results since it will stop pos-

sible action of endogenous phenolases during green

coffee solubilization. Comparison between different

extraction methods showed that the use of aqueous

methanol (40%, v/v), followed by precipitation of

colloidal material using the Carrez solutions (sol. I,

21.9 g zinc acetate and 3 ml of glacial acetic acid

dilute in distilled water up to 100 ml; sol. II, 10.6 g

potassium hexacyanoferrate (II) in 100 ml distilled

water). Reverse-phase HPLC with an ordinary C

column, gradient elution from a pH 2.5 solution of

0.01 mol l

1

tripotassium citrate or trifluoracetic acid

to methanol, increasing from 20% to 70% and UV

detection at 325 nm, has become a popular proced-

ure. The difficulties of commercial standards may be

partially overcome by the isomerization of the com-

mercially available 5-caffeoylquinic acid solution at

pH 8 with ammonia and heating in a boiling-water

bath. Using this procedure a mixture of the three

caffeoylquinic isomers is obtained. The same proced-

ure may be applied to other isolated fractions of

feruloylquinic acid to help peak assignment. Quanti-

fication may be achieved by calculation based on

extinction coefficients of different isomers found in

the literature.

Other Organic Acids and Coffee Acidity

0023Together with chlorogenic acid, other organic acids

such as citric, malic, oxalic, and tartaric acids are

found in green coffee. Other acids may be generated

tbl0003 Table 3 Content of chlorogenic acid groups in green, roasted, and instant coffees

Chlorogenic acid groups Green Arabica Medium roasted Arabica Green Robusta Medium roasted Robusta Instant coffee

(average of 13 samples)

CQA 5.8 2.0 6.8 1.8 4.3

diCQA 0.9 0.2 1.4 0.1 0.4

FQA 0.3 0.1 0.6 0.2 1.1

Total 7.0 2.3 8.8 2.1 5.8

Results expressed in g% dry basis.

CQA, caffeoylquinic acid; diCQA, di-caffeoylquinic acid; FQA, feruloylquinic acid.

Adapted from Trugo LC (1984) PhD thesis. University of Reading, UK.

COFFEE/Analysis of Coffee Products 1503

during coffee roasting, contributing to the acidity of

the beverage. Consequently, organic acids such as

lactic, pyruvic, succinic, glutaric, and others, al-

though present only in small amounts, will contribute

to the overall coffee acidity and flavor. Acidity is very

important for the taste of coffee beverage. The term

‘sour’ is frequently used to define the undesirable

taste of over-acid coffees. However, their harmonic

composition also contributes to the desirable taste of

good coffees. Coffee acidity can be assessed by pH

measurement or by titration with an alkali solution.

Undissociated molecules of acids may have specific

flavor which is not exactly related to acidity and in

some cases the analysis of individual acids may be

useful for quality control. In this case, ion chroma-

tography with electrochemical detection or gas chro-

matography after derivatization seems to be a more

adequate analytical technique. The use of electro-

chemical detection with continuous-flow injection

analysis may be useful for quality control involving

a large amount of samples.

Aroma Analysis

0024 The volatile composition of roasted coffee is ex-

tremely complex. The green coffee shows relatively

simple volatile composition which is completely

modified during roasting. Many classes of volatiles

are generated during coffee roasting (Table 4). Over a

thousand compounds have already been identified in

the volatile fraction of roasted coffee but the compos-

ition of the characteristic coffee aroma is not yet

completely established. The analysis of the head-

space composition of coffee by gas chromatography

is the most common approach. The use of mass spec-

trometry coupled to gas chromatography (GC-MS)

has proven to be very useful for structure elucidation

of the aroma composition. Sometimes it is necessary

to concentrate the volatile fraction, which may be too

diluted to be detected in the chromatogram. In those

cases codistillation procedures using the Likens–

Nickerson apparatus or entrapment in solid sorbents

such as tenax and porapack may be very useful. Some

volatiles, although present in only minute amounts,

may have very low sensory thresholds and be

extremely relevant for coffee flavor. Other useful

purge-and-trap methods may involve freezing at

very low temperatures in liquid nitrogen environment

of the coffee volatiles carried by an inert gas in capil-

lary fused silica tubes and thermal purging of the

concentrated volatiles direct to the gas chromatog-

raphy column.

0025Due to the complex gas chromatograms obtained

from volatile coffee analysis, some research has been

carried out using the sniffing port technique. In this

approach, part of the column effluent is diverted from

the detector to the sniffing person who will register all

characteristics of the compound corresponding to the

detected chromatographic peak. This sensorial evalu-

ation (aromagram) will provide very useful informa-

tion about the sensory impact of specific components.

Further chemical and structure research will then be

directed to the selected peaks which are relevant for

aroma composition. Electronic chemical gas sensor

systems, called electronic noses, have recently been

developed in order to assess the sensorial impact of

specific volatile compounds. This kind of equipment

has been applied to the analysis of food, cosmetic,

and other related materials and, although producing

some controversial results, it appears to present some

potential for coffee and other food analysis. The gas

sensors commercially available are mainly based

on metal oxide semiconductors, organic conducting

polymers, and piezoeletric crystal sensors. Sensors

based on fiberoptic, electrochemical, and bimetal

principles are due to become commercial. Statistical

methods such as principal components analysis, clus-

ter analysis, canonical discriminant analysis, artificial

neural network, and radial basis function are gener-

ally used to treat the electronic nose data.

0026Roasted coffee is considered to be one of the more

complex matrices regarding aroma composition and

has been chosen to validate some commercial instru-

ments. The instruments appeared to be capable of

differentiating between Arabica and Robusta coffee.

Sensor output and roasting level were found to pro-

duce highly correlated data submitted to another

electronic nose equipment. However, it is apparent

that there is a need for further investigation and

development in this sort of equipment for it to be

significant in coffee quality control.

Detection of Coffee Contaminants

0027The characteristic thermal conditions used for coffee

roasting at very high temperatures make the coffee

tbl0004 Table 4 Volatile compounds important for coffee aroma and

their possible precursors formed during roasting

Class of components Precursors

Pyrrole and furan Reaction products of amino acids þ

carbohydrate

Furanone Sucrose

Furan aldehyde Saccharose, arabinogalactan

Pyrazine Hydroxy amino acids

Pyridine Trigonelline, hydroxy amino acids

Phenol Chlorogenic acid

Adapted from De Maria CAB (1995) PhD thesis. Federal University of Rio

de Janeiro.

1504 COFFEE/Analysis of Coffee Products

beverage a product with very low risk of microbial

contamination. However, during harvesting, drying,

and storage, some extraneous material may be

incorporated to the product, such as in any stage of

food production. Inadequate storage conditions may

favor fungi development, producing mycotoxins

which may be resistant to coffee roasting. Ochratoxin

A appears to be the most reported mycotoxin detected

in coffee products. Although having potent nephro-

toxin activity and being immune-suppressant, terato-

genic, and carcinogenic, its eventual presence in coffee

products does not appear to represent a public health

risk due to the very low levels found (< 8mgkg

1

In these cases the use of HPLC or enzyme-linked

immunosorbent assay techniques is recommended

for analysis. HPLC analysis usually needs extensive

concentration and clean-up by means of solid-phase

extraction, with polar, nonpolar, and affinity phases,

reverse-phase chromatography, and fluorimetric de-

tection (333 nm excitation and 470 nm emission).

Atomic absorption spectrometry may be applied

for the investigation of possible contamination

with metals. HPLC and GC techniques may be indi-

cated for the analysis of polycyclic aromatic hydro-

carbons due to the use of direct fired roasters, or to

the analysis of phenols and pesticide residues due to

contaminated sacks or disinfection treatments. (See

Contamination of Food.)

Sensory Analysis

0028 Sensory analysis is still very important for coffee

classification because there is no objective approach

to define a clear correlation of chemical data with

sensorial attributes. Expert coffee tasters are still

important professionals in the coffee industry

since they produce important information on coffee

flavor directed to consumers’ expectations. The use of

sensory analysis has increased in both applied and

fundamental investigations, coupled with the use

of different statistical methods. Some statistical

methods, such as analysis of variance (ANOVA),

principal components analysis (PCA), cluster analysis

(CA) and flash analysis, have been extensively used to

assess coffee sensory attributes and consumer prefer-

ences. (See Sensory Evaluation: Sensory Characteris-

tics of Human Foods.)

Coffee Authentication

0029 Different approaches and techniques have been used

to assess coffee authenticity. The application of

unusually high processing temperatures for coffee

roasting followed by grinding to produce roasted

and ground coffee make it extremely difficult to

compare the physical and chemical characteristics of

the final product with the original green coffee beans

used. This is further complicated during elaboration

of soluble coffees where the roasted and ground

material is extracted with hot water under pressure

followed by concentration and water evaporation to

obtain the dried solubles. Chemical analysis becomes

more complex by the generation of a wide range of

pigments which interfere with sample preparation.

0030Different groups of compounds have been used as

recognition patterns to differentiate coffee species

and varieties and also adulteration by the addition

of extraneous material during processing. The ana-

lyses of specific lipids such as diterpene mono-fatty

esters, particularly cafestol and kahweol which

appear to be typical of coffee beans, have been used.

Additionally, kahweol is only found in Arabica

coffee, which may be useful for species characteriza-

tion. The fatty acid composition and also the triacyl-

glycerol profile are dependent on species and may

produce discrimating data after statistical treatment

by principal component analysis, cluster analysis,

linear discriminant analysis, or other pattern recogni-

tion method. The determination of mineral compos-

ition may be used to compare coffee varieties and also

as an aid to detect adulteration due to modification of

the typical mineral composition of coffee. The vola-

tile fraction may also be useful for coffee authentica-

tion due to the presence of typical coffee aroma from

specific compounds such as furfurylthiol and kah-

weofuran. The combination of results from volatile

and nonvolatile analyses coupled with statistical

treatments also presents potential for coffee authenti-

cation. Results from gel permeation profile, chloro-

genic acid composition, caffeine, and trigonelline

combined with head-space analysis show a potential

field for development of authentication methods. The

appearance of free sugars in instant coffees from

polysaccharide hydrolysis during industrial extrac-

tion has been explored as an approach for instant

coffee authentication. The typical presence of arabi-

nose, fructose, mannose, glucose, and galactose, with

the clear predominance of arabinose, proved to be

useful indicators of coffee authenticity. Adulteration

of coffee with other plant materials will considerably

alter the typical sugar profile. For example, the add-

ition of chicory during instant coffee manufacture

will abnormally increase the fructose content and

the use of byproducts from green coffee production

(husks, parchments, etc.) will show the appearance of

mannitol and higher levels of fructose and sometimes

xylose. After hydrolysis, pure soluble coffee is char-

acterized by high amounts of total galactose and

mannose, while the product adulterated with coffee

husks and parchments will present high total glucose

COFFEE/Analysis of Coffee Products 1505