Food Chemistry 145 (2014) 168–172

Contents lists available at ScienceDirect

Food Chemistry journal homepage: www.elsevier.com/locate/foodchem

Effect of vacuum roasting on acrylamide formation and reduction in coffee beans Monica Anese a,⇑, Maria Cristina Nicoli a, Giancarlo Verardo b, Marina Munari a, Giorgio Mirolo a, Renzo Bortolomeazzi a a b

Dipartimento di Scienze Degli Alimenti, University of Udine, via Sondrio 2/A, 33100 Udine, Italy Dipartimento di Chimica, Fisica e Ambiente, University of Udine, via del Cotonificio 108, 33100 Udine, Italy

a r t i c l e

i n f o

Article history: Received 9 May 2013 Received in revised form 23 July 2013 Accepted 12 August 2013 Available online 20 August 2013 Keywords: Acrylamide Coffee Roasting Vacuum treatment

a b s t r a c t Coffea arabica beans were roasted in an oven at 200 °C for increasing lengths of time under vacuum (i.e. 0.15 kPa). The samples were then analysed for colour, weight loss, acrylamide concentration and sensory properties. Data were compared with those obtained from coffee roasted at atmospheric pressure (i.e. conventional roasting), as well as at atmospheric pressure for 10 min followed by vacuum treatment (0.15 kPa; i.e. conventional-vacuum roasting). To compare the different treatments, weight loss, colour and acrylamide changes were expressed as a function of the thermal effect received by the coffee beans during the different roasting processes. Vacuum-processed coffee with medium roast degree had approximately 50% less acrylamide than its conventionally roasted counterpart. It was inferred that the low pressure generated inside the oven during the vacuum process exerted a stripping effect preventing acrylamide from being accumulated. Vacuum-processed coffee showed similar colour and sensory properties to conventionally roasted coffee. Ó 2013 Elsevier Ltd. All rights reserved.

1. Introduction As known, the suspected carcinogen acrylamide can form in a wide range of cooked foods, including potato derivatives, bakery products and roasted coffee (Claeys, De Vleeschouer, & Hendrickx, 2005; Friedman & Levin, 2008). The main route of acrylamide formation is represented by the Maillard reaction, which involves the reaction of asparagine with a carbonyl compound (Mottram, Wedzicha, & Dodson, 2002; Stadler et al., 2002). Due to its toxicity, efforts have been carried out to find possible technological interventions to reduce acrylamide levels in foods and thus consumer exposure. These include pre-treatments as well as formulation and/or process changes (Food Drink Europe, 2011). The application of low-pressure treatments to reduce acrylamide levels has been also explored. Vacuum can be used to remove acrylamide or to prevent its formation. In the former case, vacuum is applied to the finished product after the cooking process has been completed, in order to remove the already formed molecule. In principle, according to this approach, acrylamide can be removed by exploiting its physicochemical properties (Budavari, O’Neil, Smith, Heckelman, & Kinneary, 1996). This strategy allowed the significant removal of acrylamide from biscuits and potato chips previously hydrated at a high water content (Anese, Suman, ⇑ Corresponding author. Tel.: +39 0432558153; fax: +39 0432558130. E-mail address: [email protected] (M. Anese). 0308-8146/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.foodchem.2013.08.047

& Nicoli, 2010). Vacuum can be also applied during the heating process to minimise acrylamide formation. Vacuum frying effectively reduced (up to 94%) acrylamide formation in potato chips without affecting the colour and texture attributes, compared with samples fried under atmospheric conditions (Granda & Moreira, 2005; Granda, Moreira, & Tichy, 2004). The authors attributed this effect to the much lower temperatures used during vacuum frying. Moreover, it was observed that this technology allowed oil uptake to be greatly reduced in fried snack foods and nutrients, such as ascorbic acid and carotenoids, to be better preserved, compared to atmospheric frying (Dueik & Bouchon, 2011; Sobukola, Dueik, Munoz, & Bouchon, 2013). Although bakery products together with potato derivatives are the most important sources of acrylamide, coffee may markedly contribute to the total acrylamide content of the diet, mainly in North European countries, where its consumption is very high (Dybing et al., 2005; Granby & Fagt, 2004; Guenther, Anklam, Wenzl, & Stadler, 2007). In fact, coffee accounts for about 13% of total acrylamide in the diet of the whole population of The Netherlands, and for 27% and 39% in that of adults in Norway and Sweden, respectively (Dybing et al., 2005). However, until today, there are no viable strategies for minimising the acrylamide content in coffee, without adversely affecting the sensory quality of the finished product (EFSA, 2010; Guenther et al., 2007; Lantz, Ternité, Wilkens, H önicke, Guenther, & Van der Stegen, 2006).

M. Anese et al. / Food Chemistry 145 (2014) 168–172

The aim of this work was to investigate the effect of vacuum roasting on acrylamide formation and reduction in coffee beans. To this purpose, low-pressure treatments combined or not with roasting at atmospheric pressure were applied, in order to achieve roasted beans comparable in terms of weight loss, residual moisture and colour. In order to compare the effects of the different roasting processes, data were expressed as a function of the thermal effect F, which represents the time–temperature combination received by coffee beans at each roasting time. The effect of the roasting processes on coffee sensory properties was also evaluated. 2. Materials and methods 2.1. Coffee beans Green coffee beans of Coffea arabica with moisture content of 7.60 ± 0.01% by weight were used. They had length, width and depth mean values (n = 20) of 8.6 ± 0.6, 7.0 ± 0.3 and 3.4 ± 0.2 mm, respectively, and a sphericity of 0.01824 ± 0.00023 (Severa, Buchar, & Nedomovà, 2012). 2.2. Roasting Experiments were conducted by using an apparatus consisting of an oven (5Pascal, VS-25 SC, Trezzano S/N, Milano, Italy), equipped with heated plates for optimal heat transfer under vacuum conditions, and connected to a vacuum pump. Roasting was carried out for increasing lengths of time at atmospheric pressure (hereafter called conventional roasting), at atmospheric pressure for 10 min followed by vacuum treatment (0.15 kPa; hereafter called combined conventional-vacuum roasting), or under vacuum (i.e. 0.15 kPa; hereafter called vacuum roasting). Once the desired temperature was reached (i.e. 200 ± 1 °C), weighed aluminium dishes containing approximately 10 g of green coffee beans were introduced in the geometrical centre of the oven on a heated plate and the vacuum pump was immediately switched on. The time needed to achieve the desired vacuum was less than 10 s. Computation of treatment duration started once the set pressure value was achieved. After the treatments, samples were immediately removed from the oven and cooled to room temperature. Afterwards they were transferred to plastic vessels with pressure lids and stored at 18 °C until analyses were performed. In all cases, the time between the end of the vacuum treatment and analytical determinations never exceeded 24 h. 2.3. Analysis of acrylamide Acrylamide determination was performed according to the method of Bortolomeazzi, Munari, Anese, and Verardo (2012). In brief, acrylamide was extracted by 10 mL of water and the extract purified by a single SPE column consisting of 0.5 g of a mixture of C18, strong cation (SCX) and anion exchange (SAX) sorbents in the ratio 2/1.5/1.5 (w/w/w). The quantitation was carried out by liquid chromatography-tandem mass spectrometry using d3-acrylamide as internal standard. The relative response factor of acrylamide with respect to d3-acrylamide was calculated daily as the average of the response factors obtained by analysing a standard solution a minimum of three times. 2.4. Determination of total solid content Total solid content was determined by gravimetric method (AOAC, 1995).

169

2.5. Weight loss Sample weight roast loss (WL) was calculated as the percentage weight difference between the initial and final weights of the roasted sample. 2.6. Colour analysis Colour analysis was carried out on ground sample using a tristimulus colorimeter (Chromameter-2 Reflectance, Minolta, Osaka, Japan) equipped with a CR-300 measuring head. The instrument was standardised against a white tile before measurements. Colour was expressed in L⁄, a⁄ and b⁄ scale parameters and a⁄ and b⁄ were used to compute the hue angle (tan1 b⁄a⁄) (Clydesdale, 1978). 2.7. Temperature monitoring and thermal effect determination Temperature changes of coffee during roasting were measured by a copper-constantan thermocouple probe (Ellab A/S, Hilleroed, Denmark), whose tip (2.0 mm) was placed on the coffee bean surface. The thermal effect F (min) was computed using the following equation (Ball, 1923):

F¼

Z

t

10ðTT r Þ=z  dt

ð1Þ

0

where Tr is the reference temperature, which was chosen equal to 200 °C, roasting processes being generally carried out at temperatures around 200 °C (Clarke, 1987), T is the actual temperature of the treatment (°C), t is the time (min) of the treatment, and z represents the increase in temperature that causes a 10-fold increase in the reaction rate, which was reported to be equal to 56 °C for the browning reaction of coffee (Sacchetti, Di Mattia, Pittia, & Mastrocola, 2009). 2.8. Sensory analysis The procedure described by Manzocco and Lagazio (2009) was followed. A panel of twelve Italian assessors was selected. Judges were usual coffee consumers, aged between 20 and 60 years and approximately balanced between males and females. They all had a minimum of 2 years of experience in discrimination and descriptive sensory methods. For sensory testing, 5 g of coffee powder were served in 50-mL capacity odourless plastic cups at ambient temperature. Coffee samples were indicated by a three-digit code and submitted to the panel paired with a reference sample (i.e. the conventionally roasted coffee powder). Assessors were asked to sniff the samples after the reference one and evaluate the intensity of odour, differentiating the treated sample from the reference sample on a 9-cm unstructured scale anchored with ‘‘reference’’ corresponding to the highest odour intensity. Due to coffee persistent flavour, only one sample was evaluated at each session and assessors evaluated the samples twice at different sessions. 2.9. Image acquisition Coffee powder images were acquired by using an image acquisition cabinet (Immagini & Computer, Bareggio, Italy) equipped with a digital camera (EOS 550D, Canon, Milano, Italy). In particular, the digital camera was placed on an adjustable stand positioned 60 cm above a black cardboard base where the Petri dish containing the sample was placed. Light was provided by 4 100W frosted photographic floodlights, in a position allowing minimum shadow and glare. Images were saved in jpeg format resulting in 3456  2304 pixels.

170

M. Anese et al. / Food Chemistry 145 (2014) 168–172

2.10. Statistical analysis

80

(a)

convenonal

70

Lightness (L*)

Analyses were carried out at least twice in two replicated experiments; therefore each value is the average of at least four analyses. Coefficients of variation, expressed as the percentage ratio between the standard deviations and the mean values, were lower than 10 for acrylamide, weight loss and colour values, and 0.5 for total solid content. Analysis of variance was carried out and differences among means were assessed by using the Tukey or Student’s t test (STATISTICA for Windows, 5.1, Statsoft Inc., Cary, NC). Means were considered significantly different at p < 0.05.

combined

60

vacuum

50 40

30 20 10 0 0

5

10

3. Results and discussion

15

20

25

30

20

25

30

F(min)

200

80

Hue angle (tan-1 b*/a*)

Fig. 1 shows the temperature changes of coffee beans during conventional, combined conventional-vacuum and vacuum roasting. The temperature monitored during the vacuum treatment was lower than that recorded during the other two processes, especially in the initial stages of roasting. This can be explained by the higher rate of water vaporisation when compared with the process carried out at atmospheric pressure, due to the very low pressure generated inside the oven. As a consequence, different thermal effects (F) were obtained at the same length of roasting among the considered processes (Table 1). The degree of roast of coffee beans during conventional, combined conventional-vacuum, or vacuum roasting processes was evaluated by means of the weight loss determination (Table 1). As expected, in all cases, in the early stages of roasting coffee weight loss proceeded faster than in the later stages and was mainly due to water evaporation. At the beginning of the

(b)

70 60 50

40 30 20 10

0 0

5

10

15

F(min) Fig. 2. Lightness (a) and hue angle (b) values of coffee subjected to conventional, combined conventional-vacuum and vacuum roasting as a function of the thermal effect (F).

180

Temperature ( °C)

160 140 120

100 convenonal combined vacuum

80 60 40 20 0 0

10

20

30

40

50

60

Time (min) Fig. 1. Temperature changes during conventional, combined conventional-vacuum and vacuum roasting processes.

treatment, for a constant F value, weight losses during the vacuum roasting process were significantly higher than those occurred during the conventional and combined conventional-vacuum roasting. This can be attributable to a lower resistance of water to escape, generated by the very low pressure inside the oven. From Table 1 it can be noted that the coffee beans subjected to the conventional, combined conventional-vacuum and vacuum processes reached comparable weight losses at a similar F value of approximately 20 min, which was attained at 30, 40 and 45 min roasting, respectively. According to the literature (Clarke, 1987), this weight loss value corresponded to a dark degree of roast. It is noteworthy that a medium degree of roast, accounting for weight losses ranging from 6% to 8% (Clarke, 1987), was achieved by applying F values of approximately 3.9 and 4.2 min

Table 1 Thermal effect (F) and weight loss (WL) received by coffee beans during conventional, combined conventional-vacuum and vacuum roasting processes. Roasting time (min)

3 5 7 10 15 20 30 40 45 60

Conventional roasting

Combined conventional-vacuum roasting

Vacuum roasting

F (min)

WL (%)

F (min)

WL (%)

F (min)

WL (%)

0.9 ± 0.3 1.5 ± 0.4 2.4 ± 0.5 3.9 ± 0.5 8.1 ± 0.7 11.2 ± 0.4 18.2 ± 0.7

2.5 ± 0.1 4.2 ± 1.7 5.0 ± 0.5 6.8 ± 0.1 10.4 ± 0.1 12.3 ± 0.8 13.7 ± 0.2

0.7 ± 0.2 1.5 ± 0.5 2.5 ± 0.7 4.2 ± 0.8 7.0 ± 1.1 10.0 ± 1.3 16.0 ± 1.6 19.9 ± 3.7

2.5 ± 0.1 4.2 ± 1.7 5.0 ± 0.5 6.8 ± 0.1 10.7 ± 1.0 11.7 ± 0.5 12.9 ± 0.7 13.4 ± 0.6

0.3 ± 0.1 0.5 ± 0.2 1.1 ± 0.3 1.6 ± 0.6 3.8 ± 0.7 6.9 ± 0.6 12.1 ± 0.9 17.3 ± 1.2 19.8 ± 1.3 28.9 ± 1.0

1.7 ± 0.8 2.6 ± 0.5 4.3 ± 0.5 5.7 ± 0.8 8.0 ± 0.1 9.5 ± 0.5 12.7 ± 0.5 13.1 ± 0.7 13.3 ± 0.2 14.2 ± 0.9

Data are the mean of at least two repetitions made on two replicated sets of experiments ± sd.

171

M. Anese et al. / Food Chemistry 145 (2014) 168–172

Table 2 Images of ground coffee subjected to conventional, combined conventional-vacuum and vacuum roasting up to medium and dark degrees of roast. The corresponding thermal effects (F) are also reported.

Degree of roast

Vacuum roasting

Conventional

Combined

roasting

conventionalvacuum roasting

medium

F=3.9 min

F=4.2 min

F=3.8 min

F=18.2 min

F=19.9 min

F=19.8 min

dark

900

convenonal

800

combined

8

700

vacuum

600 500 400

300 200

0 5

10

15

20

25

a

a’

6 4 2

0

100 0

a

b’ Odor intensity

Acrylamide (ng/gdm)

conventional vacuum

10

1000

30

F(min) Fig. 3. Acrylamide concentration of coffee beans subjected to conventional, combined conventional-vacuum and vacuum roasting as a function of the thermal effect (F).

for the conventional and combined conventional-vacuum processes respectively, and 1.6–3.8 min for the vacuum process (Table 1). Roasting of coffee under different pressure conditions was also responsible for different rates of browning. Fig. 2 shows the lightness and hue angle changes of coffee beans subjected to conventional, combined conventional-vacuum and vacuum roasting processes as a function of the F value. As known, a decrease in these two colour parameters is an index of browning development due to the formation of melanoidins (i.e. the brown polymers that are formed during the final stage of the Maillard reaction). Only

medium-roasted

dark-roasted

Fig. 4. Odour intensity of coffee powders medium or dark roasted under conventional (reference) and vacuum conditions. Within each degree of roast, different letters indicate significant difference (p < 0.05).

slight differences in lightness (Fig. 2a) were observed among the different samples. By contrast, the decrease in hue angle (Fig. 2b) was greater in the vacuum-processed coffee up to an F value of approximately 3.8 min. At increasing F values, browning was significantly higher for the conventionally roasted coffee. Nevertheless, these differences in hue angle were almost not appreciable by visual observation of the medium and dark roasted ground samples (Table 2). Fig. 3 shows the changes of acrylamide concentration in coffee beans subjected to conventional, combined conventional-vacuum and vacuum roasting as a function of F. As is well known, the amount of acrylamide increases exponentially with roasting time, reaches a maximum and then decreases rapidly (Bagdonaite, Derler, & Murkovic, 2008; Guenther et al., 2007; Lantz et al., 2006; Taeymans et al., 2004; Sßenyuva & Gökmen, 2005). Such a decrease

172

M. Anese et al. / Food Chemistry 145 (2014) 168–172

occurs when the rate of degradation exceeds the rate of formation, probably due to the reaction of acrylamide with other reactive species present in the coffee (Biedermann, Biedermann-Brem, Noti, & Grob, 2002). Recently, Pastoriza, Rufián-Henares, and Morales (2012) stated that acrylamide may react with coffee melanoidins during roasting, leading to a net decrease in acrylamide content. In our experimental conditions, differences in acrylamide formation and reduction among the roasting processes considered were observed (Fig. 3). In particular, acrylamide concentration in the vacuum-roasted coffee sharply reached a maximum at an F value of 1.6 min and then decreased. In the case of the conventional and combined conventional-vacuum roasting, a maximum acrylamide formation was attained at F values ranging from 1.5 to 4.2 min. No significant differences in acrylamide concentration among the coffee beans roasted according to the different processes were found at F values higher than 15 min. These differences in acrylamide reduction of coffee beans subjected to roasting under the different pressure conditions adopted can be attributable to a stripping effect exerted by the low pressure generated inside the oven during the vacuum process towards water as well as water-soluble low-molecular-weight compounds, including volatile compounds and acrylamide. Such a stripping effect would prevent acrylamide from being accumulated. In order to study the effects of roasting under low pressure conditions on coffee quality, sensory analysis was carried out by sniffing ground coffee subjected to conventional, combined conventionalvacuum and vacuum roasting with weight losses of about 7.0% and 13.5%, accounting for medium and dark degrees of roast, respectively (Fig. 4). It can be observed that the medium-roasted coffee samples obtained by means of the conventional and vacuum processes and having different acrylamide levels, were not perceived as different by the assessors. On the contrary, the vacuum darkroasted coffee was judged to present a slightly but significantly lower odour intensity with respect to the reference (conventionally roasted) sample. No significant differences were found between the dark-roasted coffees subjected to the conventional and combined conventional-vacuum processes (data not shown). 4. Conclusion Roasting under low pressure conditions gave medium-roasted coffee with 50% lower acrylamide levels, as compared to the conventionally roasted coffee, with minimal impact on the sensory properties. Although further research should be conducted at pilot and industrial scale to find optimum process conditions, these results suggest that vacuum roasting could have a great economic impact, as medium-roasted coffee generally has higher acrylamide levels than dark-roasted coffee. Indeed, from the industrial point of view, safety could be a new factor influencing coffee commercial opportunities, especially for the American and North European markets, where medium-roasted coffee consumption, and thus acrylamide intake, are relatively high. References A.O.A.C. Official Method 925.09 (1995). Official methods of analysis of AOAC international (16th ed.). WV, USA: Arlington. Anese, M., Suman, M., & Nicoli, M. C. (2010). Acrylamide removal from heated foods. Food Chemistry, 119, 791–794.

Bagdonaite, K., Derler, K., & Murkovic, M. (2008). Determination of acrylamide during roasting of coffee. Journal of Agriculture and Food Chemistry, 56, 6081–6086. Ball, C. O. (1923). Thermal process time for canned food. Bull National Resources Council, 7, 37. Biedermann, M., Biedermann-Brem, S., Noti, A., & Grob, K. (2002). Methods for determining the potential of acrylamide formation and its elimination in raw materials for food preparation, such as potatoes. Mitteilungen aus Lebensmitteluntersuchung und Hygiene, 93, 653–667. Bortolomeazzi, B., Munari, M., Anese, M., & Verardo, G. (2012). Rapid mixed mode solid phase extraction method for the determination of acrylamide in roasted coffee by HPLC-MS/MS. Food Chemistry, 135, 2687–2693. Budavari, S., O’Neil, M. J., & Smith, A. (1996). The Merck Index. Whitehouse Station, NJ: The Merck Index & Co Inc.. Claeys, W. L., de Vleeschouer, K., & Hendrickx, M. E. (2005). Quantifying the formation of carcinogens during food processing: acrylamide. Trends in Food Science & Technology, 16, 181–193. Clarke, R. J. (1987). Roasting and grinding. In R. J. Clarke & R. Macrae (Eds.). Coffee technology (Vol. 2, pp. 73–114). London: Elsevier Applied Science Publishers LTD. Clydesdale, F. M. (1978). Colorimetry-methodology and applications. Critical Reviews in Food Science and Nutrition, 10, 243–301. Dueik, V., & Bouchon, P. (2011). Vacuum frying as a route to produce novel snacks with desired quality attributes according to new health trends. Journal of Food Science, 76, E188–E195. Dybing, E., Farmer, P. B., Andersen, M., Fennell, T. R., Lalljie, S. P. D., Müller, D. J. G., et al. (2005). Human exposure and internal dose assessments of acrylamide in food. Food and Chemical Toxicology, 43, 365–410. EFSA European Food Safety Authority (2010). Results on arylamide levels in food from monitoring year 2008. EFSA J., 8(5), 1599–1630. Available from www.efsa.europa. Food Drink Europe (2011). Food drink europe acrylamide toolbox. . Available from www.fooddrinkeurope.eu/publications/category/toolkits/. Friedman, M., & Levin, C. E. (2008). Review of methods for the reduction of dietary content and toxicity of acrylamide. Journal of Agriculture and Food Chemistry, 56, 6113–6140. Granby, K., & Fagt, S. (2004). Analysis of acrylamide in coffee and dietary exposure to acrylamide from coffee. Analytica Chimica Acta, 520, 177–182. Granda, C., & Moreira, R. G. (2005). Kinetics of acrylamide formation during traditional and vacuum frying of potato chips. Journal of Food Process Engineering, 28, 478–493. Granda, C., Moreira, R. G., & Tichy, S. E. (2004). Reduction of acrylamide formation in potato chips by low-temperature vacuum frying. Journal of Food Science, 69, 405–411. Guenther, H., Anklam, E., Wenzl, T., & Stadler, R. H. (2007). Acrylamide in coffee: review of progress in analysis, formation and level reduction. Food Additives & Contaminants, 26, 60–70. Lantz, I., Ternité, R., Wilkens, J., Hönicke, K., Guenther, H., & van der Stegen, G. H. D. (2006). Studies on acrylamide levels in roasting, storage and brewing of coffee. Molecular Nutrition & Food Research, 50, 1039–1046. Manzocco, L., & Lagazio, C. (2009). Coffee brew shelf life modelling by integration of acceptability and quality data. Food Quality and Preference, 20, 24–29. Mottram, D. S., Wedzicha, B. L., & Dodson, A. T. (2002). Acrylamide is formed in the Maillard reaction. Nature, 419, 448–449. Pastoriza, S., Rufián-Henares, J. Á., & Morales, F. J. (2012). Reactivity of acrylamide with coffee melanoidins in model systems. LWT-Food Science and Technology, 45, 198–203. Sacchetti, G., Di Mattia, C., Pittia, P., & Mastrocola, D. (2009). Effect of roasting degree, equivalent thermal effect and coffee type on the radical scavenging activity of coffee brews and their phenolic fraction. Journal of Food Engineering, 90, 74–80. S ß enyuva, H. Z., & Gökmen, V. (2005). Study of acrylamide in coffee using an improved liquid chromatography mass spectrometry method: investigation of colour changes and acrylamide formation in coffee during roasting. Food Additives & Contaminants, 22, 214–220. Severa, L., Buchar, J., & Nedomovà, S. (2012). Shape and size variability of roasted Arabica coffee beans. International Journal of Food Properties, 15, 426–437. Sobukola, O. P., Dueik, V., Munoz, L., & Bouchon, P. (2013). Comparison of vacuum and atmospheric deep-fat frying of wheat starch and gluten based snacks. Food Science and Biotechnology, 22(S), 177–182. Stadler, R. H., Blank, I., Varga, N., Robert, F., Hau, J., Guy, P. A., et al. (2002). Acrylamide from Maillard reaction products. Nature, 419, 449. Taeymans, D., Wood, J., Ashby, P., Blank, I., Studer, A., Stadler, R. H., et al. (2004). A review of acrylamide: an industry perspective on research, analysis, formation, and control. Critical Reviews in Food Science and Nutrition, 44, 323–347.