A Calorimetric Study of Self Heating in Coffee and Chicory
A Calorimetric Study of Self Heating in Coffee and Chicory
A Calorimetric Study of Self Heating in Coffee and Chicory
(1982) 17,451-460
ALOIS =MY*
AND
P. LAMBELET
Summary
The technique of heat flow calorimetry was used to determine the specific heat
of coffee and chicory products and to study their thermal behaviour above 20C.
Intensive exothermic reactions were particularly evident when measurements
were made with sealed cells. The data obtained help in the understanding of
how exothermic reactions (and self-heating) occurring in such foodstuffs can
bring them above their minimum ignition temperature.
Introduction
The principle of heat flow calorimetry is well known (Calvet & Prat, 1956;
Hemminger & Hohne, 1979); this is one of the many techniques of thermal
analysis. Instruments based on this principle have been used for many investigations of thermal properties.
Our interest in this technique lies especially in the detection and characterization of exothermic phenomena in foods and also in the possibility of determining specific heat values above 20C. Calorimetric information is important
for designing or operating heating systems or machines. A better knowledge of
these food properties should also help to understand how certain food products
can be brought above their minimum ignition temperature and can ignite
spontaneously.
Materials and methods
Instrument
The instrument we have used mostly for these measurements is a heat flow
calorimeter (Setaram C80) of Calvet type. A schematic representation of the
Authors address: Research Department, Nestle Products Technical Assistance Co. Ltd,
452
Calorimeter
Setaram C80 calorimeter is shown in Fig. 1. This instrument allows the analysis
of relatively large quantities of non-homogeneous material ( 2 . M . O g whole
coffee beans). The temperature range lies between ambient and 300C. The
samples are heated very slowly (generally at 1C min-I).
Specific heat measurements were performed in very thin-walled cells. For
studying the thermal behaviour of the food products at high temperatures,
different types of thick-walled cells, which support a pressure rise, were used.
Calorimetric diagrams obtained with sealed cells were similar to those obtained
when the samples were held under a constant pressure of 20-25 bar inert gas
during heating; the reasons why are explained in detail elsewhere (Raemy,
1981).
For the measurements of coffee and chicory presented here, the samples were
heated in sealed cells, the use of which is particularly easy; the products are
analysed in presence of a fixed amount of oxygen.
453
Time
reproduced very accurately. This allows the measurement of single values of the
specific heat, for example at 30C, by increasing the temperature from T I = 30C
to T2 = 33C. Two measurements (empty cell and sample) or three (empty cell,
sample and standard) are necessary to determine the specific heat of a substance
(Fig. 2). The reference cell is an empty piece of metal, built in such a way that all
two or three curves produced have identical shapes.
One measurement of the sample and one of the empty cell allow the determination of the unknown specific heat value, at a fixed temperature. As A T is
the same in both measurements, the following formula is obtained,
where C2 is the specific heat, under constant pressure, of the product, QOis the
required heat quantity (empty cell measurement), Qz is the required heat
quantity (product measurement), m2 is the mass of product and A T is the
temperature increase from TI to 7;.
The heat quantity values corresponding to the surface between both equilibrium temperatures must be introduced in joules (or calories); it is therefore
necessary to know the sensitivity curve (Fig. 3) of the calorimeter, determined
by Joule effect.
454
I00
200
Temperature ("C)
300
Sample description
Humidity
(% wt)
C
(J g-l "C-l)
C
(calg-l O C - ' )
7.5
1.85
0.44
4.5
1.46
0.35
1.59
0.38
2.5
1.46
0.35
2.6
2.4
2.2
1.46
1.42
1.42
0.35
0.34
0.34
3.5
1.67
0.40
4.5
1.34
0.32
2.5
1.46
0.35
455
Table 1 gives some specific heat values of coffee and chicory products
obtained at 30C: values between 1.2 J g-l "C-l (or 0.3 cal g-' "C-') and
2.0 J g-' "C-' (or 0.5 cal g-' "C-l) were generally found for dry materials. For
products like green coffee beans or soluble powders, small moisture differences
have an appreciable influence on their specific heat values.
In a restricted temperature range, an increase of the specific heat values of
solid foods is generally observed when temperature increases, as long as no
physical or chemical changes occur. This can be controlled continuously by
heating substance and empty cell at a constant rate: the deviation between both
heat flow curves is proportional to the specific heat of the sample. Thus,
c
2
=
RZZ
A Uz
(d T/dt) S '
(3)
where AUz is the calorimeter signal (between product and empty cell
measurements), dT/dt is the heating rate and S is the calorimetric sensitivity.
By performing a measurement of a standard as well, an equation similar to
eqn (2) gives the unknown specific heat directly by comparing heights. Figure 4
shows such curves between 20 and 95C; the specific heat of green Arabica
coffee beans increases from 1.85 to about 2.4 J g-' "C-' ;this represents a rate of
about 30%.
For green coffee beans or coffee products there is also a weak endothermic
phenomenon at about 60C (fusion of one component?). It is clear that for food
products which are in the form of fine powders, a more conventional differential
scanning calorimeter like Mettler TA 2000B can also be used; methods are not
very different and the results may even be obtained more quickly.
For liquid coffee or chicory extracts, specific heat values are much higher than
for powders, due to the large amount of water. Following approximately the
law of proportional additivity or Siebel's rule (Mohsenin, 1980), the
Ti me
456
specific heat value, at 30"C, is about 3.0 J g-l "C-' (0.7 cal g-' "C-I) for a concentrate extract (50% powder and 50% water) and approaches to 4.0 J g-' "C-'
(1.0 cal g-' "C-I) for very dilute solutions.
Specific heat values of liquid coffee extracts (Riedel, 1974) and of other foods
(Riedel, 1977;Narain et al., 1978; Putranon, Bowrey & Fowler, 1980; Hwang &
Hayakawa, 1979; Levy, 1979; Keppeler, 1979) can be found in the literature:
they were most often determined by calorimetric techniques. General compilation of thermal properties of foods and the methods used for their determination is given by Polley (1980) and Mohsenin (1980).
Results
Exothermic reactions
Samples of green coffee beans and of ground dried chicory roots were
analysed by heating at constant rates, generally 1C min-I. The purpose of these
measurements was to detect exothermic reactions, due to pyrolysis reactions.
With open cells at atmospheric pressure, the calorimetric curves showed an
endothermic peak above 100C (water phase transition) and at higher temperatures, an exothermic tendency. Measurements performed either under
20-25 bar pressure of inert gas (Raemy, 1981) or with sealed cells showed
similar intensive exothermic phenomena. These reactions began at about 140C
for green coffee beans; the temperature range where they occurred was even
lower for dried chicory. The measured enthalpies had values around 250-420 J
g-' (60-100 cal g-I) for green coffee beans and values as high as 540-640 J g-'
(130-150 cal g-I) for dried chicory. These exothermic phenomena correspond to
roasting and carbonization of the products. Figure 5 presents the calorimetric
curves of green Arabica coffee beans from Mexico and dried chicory from
Austria.
Exothermic peaks were still evident in calorimetric diagrams of roasted coffee
Dried chicory
I"C/rnin
Exothermic
Exothermic
Coffee beans
50
100
150
200
250
Temperature ("CI
Figure 5. Calorimetric curves of dried chicory and green coffee beans (both heated in
sealed cells).
457
Coffee io/o)
Temperature
(OC)
Figure 6. Calorimetric curves of soluble coffee and chicory powders (heated in sealed
cells).
or chicory, and even those of soluble powders containing coffee, chicory or both
components. Soluble powders (from pilot plant batches) with different percentages of coffee and chicory were compared (Fig. 6). When both components
were present in the powder, the exothermic peak was a poorly resolved doublet;
the first or left part is due primarily to chicory, the second to coffee. Enthalpies
as high as 540-640 J g-' (130-150 cal g-') were stili measured for powders
containing only chicory.
These exothermic phenomena are mainly attributed to the carbohydrates
contained in both products. It is generally agreed that green coffee beans contain
about 55% carbohydrate and that chicory contains about 75%. These values are
little changed after roasting. Some of the carbohydrate is water soluble (about
40% of the total carbohydrate in coffee and 70% in chicory); this explains why
exothermic reactions are still detected for the corresponding soluble powders.
The remaining carbohydrate (60% for coffee) is found in the spent grounds.
Figure 7 presents the calorimetric curves of a soluble coffee powder and of spent
grounds, showing the highest peak for spent grounds. Such self-heating properties were also found with other foods having high contents of carbohydrate, e.g.
most cereals. Figure 8 shows the calorimetric diagrams of wheat (variety:
Scheppers) and whole rice (variety: Camolino).
458
IC/min
I50
I20
200
Temperature ("C)
250
Figure 7. Calorimetric curves of a soluble coffee powder and of spent grounds (both
heated in sealed cells).
IC/min
I00
I50
200
2 50
Temperature ( " C )
Figure 8. Calorimetric curves of wheat and whole rice (both heated in sealed cells).
459
Q 640 - 320C.
A T = -=-mC 12
Even if a part of the heat is lost, because of other phenomena or non-adiabatic
conditions, it seems possible that some of the powder can be brought above its
minimum ignition temperature (Forsyth, 1980; Palmer, 1973) and that it will
eventually ignite,
In the presence of water, a reduced heating rate can arise for three reasons:
(1) the specific heat of the product is higher than in the absence of water; (2) heat
is absorbed by water evaporation; and (3) in the case of an already smouldering
fire, the reduction in dispersability, caused by the water, hinders flame
propagation.
It is evident that more sophisticated models (Bowes, 1976) can be developed
and that adiabatic calorimetry techniques (Townsend & Tou, 1980; Hub, 1981)
could better show how, and how far, temperature will rise in cases of spontaneous heating (thermal runaway). These alternative methods would bring us
interesting complementary data.
Conclusion
The analysis of coffee and chicory products by heat flow calorimetry has given
information on the thermal behaviour of these foods. Such knowledge of food
properties allows a better understanding of spontaneous heating, particularly in
which temperature range this phenomenon can occur and to which consequences it can lead. This is of great importance for many high-temperature
processing operations.
Acknowledgments
The authors are grateful to J . Loeliger and C. Desarzens for discussions, to Miss
P. Clerc and J. P. Marquet for helping to perform the measurements and to
R. F. Hurrell for improving their English.
460
References
Bowes, P.C. (1976) Inst. Chem. Eng. Symposium Series No. 47.
Calvet, E. & Prat, H. (1956) Microcalorimktrie. Masson et Cie, Paris.
Forsyth, V.G. (1980) Fire Prevention No. 135, p.25.
Hemminger, W. & Hohne, G. (1979) Grundlagen der Kalorimetrie. Verlag Chemie, Weinheim.
Hub, L. (1981) Instit. Chem. Eng. Symposium Series No. 68,3.
Hwang, M.P. & Hayakawa, K.I. (1979) J . Fd Sci. 44,435.
Keppeler, R.A. (1979) J . Dairy Sci. 62, 1340.
Levy, F.L. (1979) J. Fd Technol. 14,549.
Mohsenin, N.N. (1980) Thermal Properties of Foods and Agricultural Materials. Gordon and
Breach, New York.
Narain, M . , Bose Siripurapu, S.C., Jha, M. & Dwivedi, V.K. (1978) J . Fd Sci. Technol., 15,18.
Palmer, K.N. (1973) Dust Explosions and Fires. Chapman and Hall, London.
Polley, S.L., Snyder, O.P. & Kotnour, P. (1980) Fd Technol. 34,76.
Putranon, R., Bowrey, R.G. &Fowler, R.T. (1980) Fd Tech. Aust. 32,56.
Raemy, A. (1981) ThermochimicaActu, 43,229.
Riedel, L. (1974) Chem. Mikrobiol. Technol. Lebensm. 4, 108.
Riedel, L. (1977) Chem. Mikrobiol. Technol. Lebensm. 5,118.
Townsend, D.I. & Tou, J.C. (1980) ThermochimicaActa, 37,l.