3 basic chemistry and process conditions for reaction ... · 3 basic chemistry and process...

P1: SFK/UKS P2: SFKc03 BLBK221-Taylor/Linforth November 17, 2009 14:4 Printer Name: Yet to Come

3 Basic chemistry and process conditionsfor reaction flavours with particularfocus on Maillard-type reactions

Josef Kerler, Chris Winkel, Tomas Davidek and Imre Blank

3.1 INTRODUCTION

Maillard reaction technology is used by the flavour and food industry for the productionof process/reaction flavours or generating flavour upon food processing (in-process flavourgeneration). Process flavours are complex building blocks that provide similar aroma andtaste properties to those found in thermally treated foodstuffs such as meat, chocolate,coffee, caramel, popcorn and bread. The Maillard reaction between a reducing sugar anda food-grade nitrogen source is the principal underlying reaction, which is responsiblefor flavour and colour development. However, Maillard-type reactions may also give riseto undesirable molecules that need to be limited using mitigation concepts. This reviewprovides a summary of general aspects of the Maillard reaction in flavour formation in viewof reinforcing distinct desirable flavour notes. An overview of important aroma compoundsof thermally treated foodstuffs and process flavours is given. In addition, the patent literatureand other publications relating to reaction flavour production and their process conditionsare discussed. This chapter focuses on Maillard-type reactions and only partially deals withother reactions occurring during process flavour preparation (e.g. lipid oxidation).

3.2 GENERAL ASPECTS OF THE MAILLARDREACTION CASCADE

The ‘Maillard reaction’ is of great importance for flavour and colour formation of thermallytreated foodstuffs. It is a complex cascade of many different types of reactions rather thanone single reaction type, even though it is initiated by an amino-carbonyl reaction step. Thethermal generation of flavours in foods, process flavours and model systems has accordinglybeen the subject of many symposia and reviews (e.g. Parliment et al., 1989, 1994; Weenenet al., 1997; Reineccius, 1998; Tressl and Rewicki, 1999; Cerny, 2007; Yeretzian et al.,2007). A first milestone in the history of Maillard chemistry was the publication of the well-known Hodge scheme (Hodge, 1953). Although Hodge’s studies focused only on Maillardbrowning, this scheme provided a framework that also covered important reaction routes forthe formation of aroma compounds. The work of Hodge triggered a large number of studies onthe elucidation of important intermediates and pathways of the Maillard reaction (see reviewsby Ledl and Schleicher, 1990; Tressl and Rewicki, 1999). Isotopic labelling of sugars and/oramino acids in conjunction with GC-MS (gas chromatography–mass spectrometry) analysis(Tressl et al., 1993; Gi and Baltes, 1995; Keyhani and Yaylayan, 1996) and trapping of

Food Flavour Technology: Second Edition Edited by Andrew J. Taylor and Robert S.T. Linforth

© 2010 Blackwell Publishing Ltd. ISBN: 978-1-405-18543-1


52 Food Flavour Technology

reactive intermediates (Nedvidek et al., 1992; Hofmann, 1999) are key techniques for theimproved understanding of Maillard reaction pathways.

Improvements and refinements of the Hodge scheme were presented by Tressl et al.(1995). Their scheme provides an excellent overview of Maillard reaction pathways leadingto the formation of volatile compounds. For the optimisation of reaction flavours, however, astrong emphasis is required on those routes that are involved in the generation of key aromacompounds. This can be achieved by first evaluating the character-impact compounds of aprocess flavour or model system using a combination of sensorial and instrumental analysison the basis of the odour activity value concept (Grosch, 1994). The mechanistic studies canthen be focused on the key substances only. The work of Hofmann (1995) is an excellentexample of such an approach.

Figure 3.1 gives an overview of the pathways that are involved in the formation of im-portant aroma compounds during Maillard reaction. There are three main routes involved inflavour generation. All three routes start with imine formation between a reducing sugar andan amino acid. The Amadori (derived from aldoses) or Heyns (derived from ketoses) rear-rangement products are important intermediates of the early phase of the Maillard reaction.Route A (see also Section 3.2.1) leads to the formation of 1- and 3-deoxyosones, which oncyclisation, reduction, dehydration and/or reaction with hydrogen sulfide result in hetero-cyclic aroma compounds. Route B (see also Section 3.2.2) is characterised by fragmentationof the sugar chain through retro-aldolisation or �-/�-cleavage. By aldol-condensation oftwo sugar fragments or a sugar fragment and an amino acid fragment, heterocyclic aromacompounds are generated on cyclisation, dehydration and/or oxidation reactions. Alterna-tively, the fragments can react with hydrogen sulfide and form very potent alicyclic flavoursubstances. Route C (see also Section 3.3) involves the so-called Strecker degradation ofamino acids, which is catalysed by dicarbonyl or hydroxycarbonyl compounds. The reaction

Amino acid

Sugar

Amadori/Heynsrearrangement compounds

1- and 3-Deoxyosones

Cyclisation products

Amino acid

Fragmentation products

Condensation products

Cyclisation productsStrecker degradation

Retro-aldol orα (β)-cleavage

Flavour substances

+

- amino acid

- H S2reduction

- H S2reductiondehydration

A

C

B

H S, NH2 3

H S, NH , amino2 3acid fragments

Fig. 3.1 Major pathways for the formation of flavour substances during Maillard reaction. A, B and Cdenote the three key pathways.


Basic chemistry and process conditions for reaction flavours 53

Fig. 3.2 Reaction of glucose (I) and glycine leading to the Amadori compound N-(1-deoxy-D-fructos-1-yl)glycine (V, open-chain form) and related degradation reactions (adapted from Davidek et al., 2002).

is a ‘decarboxylating transamination’ and the resulting Strecker aldehydes are potent flavourcompounds. Strecker aldehydes can also be formed directly from Amadori rearrangementproducts (ARPs) or Heyns rearrangement products (HRPs).

A more detailed, but still simplified scheme is depicted in Fig. 3.2 (Davidek et al., 2002)showing the formation of Amadori compounds, N-substituted 1-amino-1-deoxy-ketoses (V)representing an important class of Maillard intermediates (Ledl and Schleicher, 1990). Theyare formed in the initial phase of the Maillard reaction by Amadori rearrangement of thecorresponding N-glycosylamines (II), the latter obtained by condensation of amino acids andaldoses such as glucose (I) as shown in pathway A. The importance of Amadori compoundsstems from the fact that their formation as well as decomposition can be initiated under mildconditions. Thus, the formation of Amadori compounds represents a low-energy pathway ofsugar degradation. The chemistry of Amadori compounds has recently been reviewed (Yay-layan and Huyghues-Despointes, 1994). Degradation of the Amadori compound V by 1,2-enolisation (pathway B) and 2,3-enolisation (pathway D) leads to the formation of 3-deoxy-2-hexosulose (VII) and 1-deoxy-2,3-hexodiulose (XII), respectively, as already suggested by



Hodge (1953). In parallel to these pathways, other �-dicarbonyls can be formed by enolisa-tion. For example, transition metal-catalysed oxidation of 1,2-enaminol IV can lead via path-way C to osones such as glucosone (IX). Pathway E gives rise to the 1-amino-1,4-dideoxy-2,3-diulose (XIII) by elimination of the C4-OH group of the 2,3-endiol (VI). If the iminoketone(VIII) formed by oxidation of 1,2-enaminol (IV) is not hydrolysed, then Strecker aldehydescan be formed in the course of pathway C by direct oxidative degradation of the Amadoricompound and decarboxylation of X, as proposed by Hofmann and Schieberle (2000a).

The so-called carbon module labelling (CAMOLA) technique (Schieberle et al., 2003;Schieberle, 2005) has been introduced as an advanced tool to quantify the relative contri-bution of carbohydrate fragments (e.g. C1–C4 fragments derived from Route B in Fig. 3.1)in comparison with transient intermediates with intact carbon chain configuration (e.g. de-oxyosones; formed via Route A in Fig. 3.1) in the generation of Maillard-derived aromacompounds. The authors used a model system containing 1:1 mixtures of unlabelled and13C6-labelled glucose as well as unlabelled proline and then measured the labelling pattern(12C6, 13C3 and 13C6) of the caramel-like odourant 4-hydroxy-2,5-dimethyl-3(2H)-furanone(Furaneol R©) by GC-MS. The study revealed that, under dry heating conditions, Furaneolwas generated entirely via the intact sugar skeleton, whereas in aqueous solution, 63% of thefuranone was derived from the recombination of two C3-fragments. This work can be con-sidered as another milestone in the history of research on Maillard chemistry. The CAMOLAtechnique has recently also been applied for studying the formation of furan in both modelsystems and foodstuffs (Limacher et al., 2008).

3.2.1 Intermediates as flavour precursors

ARPs and HRPs are relatively stable intermediates and have been detected in various heat-processed foods (Eichner et al., 1994). Since ARPs and HRPs can easily be synthesised(Van den Ouweland and Peer, 1970; Yaylayan and Sporns, 1987), their potential as flavourprecursors has been evaluated in several studies. Doornbos et al. (1981), for example, foundthat the ARP derived from rhamnose and proline is a useful precursor for the generation ofthe potent caramel-like odourant 4-hydroxy-2,5-dimethyl-3(2H)-furanone. ARPs and HRPshave also been reported to be good precursors for Strecker aldehydes and, in the absenceof oxygen, also for 1- and 3-deoxyosones (Hofmann and Schieberle, 2000a). At higher pHvalues, ARPs and HRPs easily undergo cleavage of the carbohydrate chain, yielding fissionproducts such as 2,3-butanedione and pyruvaldehyde (Weenen and Apeldoorn, 1996).

When cysteine is heated with reducing sugars, thiazolidine carboxylic acids (TCAs) areformed instead of ARPs or HRPs (de Roos, 1992). TCAs are relatively stable in anionicform, which is probably the main reason for the inhibitory effect of cysteine in Maillardreactions, especially at higher pH. This can also explain the more efficient formation ofimportant meat sulfur compounds at low pH (Hofmann and Schieberle, 1998a). A researchdisclosure (Anonymous, 1979), however, describes the use of TCAs as precursors for meatflavours. TCAs of glyceraldehyde, fructose or xylose were reacted as such or in conjunctionwith organic acids (e.g. succinic acid, malic acid and citric acid) or fatty acids (e.g. oleic acidand linoleic acid). Using a pH of 6–7 and temperatures between 50 and 100◦C, the TCA ofglyceraldehyde and cysteine was reported to result in a beef-like aroma, whereas TCAs offructose or xylose and cysteine yielded ‘meaty/savoury’ flavours.

Other important Maillard reaction intermediates are the deoxyosones. In general,1-deoxyosones are more important flavour precursors than 3-deoxyosones, whose for-mation is favoured under neutral/slightly basic and acidic pH, respectively. Although1-deoxyglucosone has been synthesised by Ishizu et al. (1967), 1-deoxyosones are too



unstable to be used as precursors. 3-Deoxyosones, however, are more stable and are easilyobtainable from compounds such as difructoseglycine (Anet, 1960). The structure and reac-tivity of various 3-deoxyosones have been extensively studied by Weenen and Tjan (1992,1994) and Weenen et al. (1998). By using various 1H NMR and 13C NMR techniques, theauthors showed that 3-deoxypentosone and 3-deoxyglucosone consist almost exclusivelyof monocyclic and bicyclic (hemi)acetal/(hemi)ketal structures. 3-Deoxypentosones and 3-deoxyhexosones are good precursors for furfural and (5-hydroxymethyl)furfural under acidicconditions. Under basic conditions, they undergo cleavage of the carbohydrate chain and canform pyrazines in the presence of an N-source.

Hofmann and Schieberle (2000b) showed that acetylformoin, which is formed from 1-deoxyhexosone, is an effective precursor for 4-hydroxy-2,5-dimethyl-3(2H)-furanone (Fu-raneol). The amounts of Furaneol obtained from acetylformoin were significantly enhancedin the presence of reductones such as ascorbic acid or methylene reductinic acid as wellas the Strecker-active amino acid proline. The reaction between acetylformoin and pro-line also resulted in high amounts of the cracker-like odourant 6-acetyltetrahydropyridine(ACTP).

A number of articles and patents report the production of meat-like aromas by reacting4-hydroxy-5-methyl-3(2H)-furanone (norfuraneol) with hydrogen sulfide or cysteine. Vanden Ouweland and Peer (1968) were first to file a patent on the use of this precursor sys-tem to prepare 3-mercaptomethylfurans, which exhibit meat-like character. Later, severalstudies identified sulfur-containing aroma compounds derived from the reaction of norfura-neol and hydrogen sulfide (e.g. Van den Ouweland and Peer, 1975; Whitfield and Mottram,1999). The latter authors showed that this precursor system is capable of producing com-pounds such as 2-methyl-3-furanthiol (MFT) and 2/3-mercapto-3/2-pentanone in relativelyhigh amounts. These thiols are also key aroma compounds of heated meat (Kerscher andGrosch, 1998).

Shu and Ho (1989) and Zheng et al. (1997) studied the reaction of the methyl homologue 4-hydroxy-2,5-dimethyl-3(2H)-furanone (Furaneol) with hydrogen sulfide or cysteine, whichalso gave rise to meat-like aromas. Among the identified sulfur-containing volatiles, however,the methyl homologue of MFT (2,5-dimethyl-3-furanthiol) was not found in the reactionmixtures.

Unilever patented processes for the preparation of savoury flavours using the precursorsystems and reaction conditions shown in Fig. 3.3 (Turksma, 1993; Rosing and Turksma,1997). When 2,5-dimethyl-2-(2-hydroxy-3-oxo-2-butyl)-3(2H)-furanone (diacetyloligomer;R and R′ = CH3; process A in Fig. 3.3), which can be obtained by heating 2,3-butanedioneunder acidic conditions (Doornbos et al., 1991), is reacted with cysteine and hydrogen sul-fide, high amounts of 2,5-dimethyl-3-furanthiol (DMFT) are generated (Turksma, 1993).For example, almost 40% yield of DMFT was obtained after 1 hour at 120◦C using a polarorganic solvent, acidic conditions and super-atmospheric pressure (100–2500 kPa). DMFT,which was found to have a ‘meaty taste and roasted meat aroma’ was also formed from2,5-dimethyl-3(2H)-furanone. Using similar conditions as for the diacetyloligomer, Rosingand Turksma (1997) reacted 4-hydroxy-2,5-dimethyl-(2-hydroxy-3-oxo-2-butyl)-3(2H)-furanone (Fig. 3.3; R1 − R3 = CH3, R4 = acetyl; process B) with cysteine and hydrogensulfide. This process resulted in a flavouring with sweet, onion-like, meaty aroma, odoursattributed to high amounts of 2,5-dimethyl-4-mercapto-3(2H)-furanone and 2,5-dimethyl-4-mercapto-3(2H)-thiophenone. However, these two sulfur compounds were not found tosignificantly contribute to the flavour of heated meat.

Mottram et al. (1998) filed a patent on flavouring agents that serve as precursors forgenerating cooked (e.g. cooked meat) flavours in foodstuffs in situ. The authors claim



OR

O

OH

O

R'

OR

SH

OR1

O

R4

OH

R3

R2

OH

XR1 R2

SHOH

Cysteine + H2S

Glycerol90–120°C1–3 hours, 100–2500 kPa

Cysteine + H2S

Glycerol90–120°C1-3 hours, 100–2500 kPa

R, R’ = CH3 or H

R1 = CH3, C2H5 or HR2 = alkyl, C1-4 or HR3 = alkyl, C1-5R4 = organic radical, C1-6, H or O-atomsX = O or S

tnarodo niaMrosrucerP

A

B

Fig. 3.3 Precursors, reaction conditions and main aroma compounds of two savoury process flavours(according to Turksma, 1993 (A), and Rosing and Turksma, 1997 (B)).

a long list of precursor substances that are capable of developing flavour during mi-crowave cooking or conventional oven cooking with reduced cooking times. This listcomprises several sulfur compounds such as hydrogen/ammonium/sodium sulfides, cys-teine, thiamine, onion and garlic as well as ‘non-sulfur-containing post-rearrangement Mail-lard products such as furanones (e.g. 4-hydroxy-5-methyl-3(2H)-furanone, 4-hydroxy-2,5-dimethyl-3(2H)-furanone, 2-methyl-4,5-dihydro-3(2H)-furanone), pyranones (e.g. maltol,5-hydroxy-5,6-dihydromaltol), 3-deoxyglucosone, ketones (e.g. cyclotene) and aldehydes.The precursor mixtures were encapsulated or spray-dried and applied to the foodstuffsthrough dusting or inclusion prior to the heat treatment. In two other studies, a similar pre-cursor system consisting of unsaturated aldehydes and hydrogen sulfide resulted in aromablocks with deep-fried notes (Van den Ouweland, 1989; Zhang and Ho, 1989).

As shown in this chapter, meat-like flavours can be obtained from the reaction of inter-mediate precursor systems such as 4-hydroxy-5-methyl-3(2H)-furanone (norfuraneol) andcysteine. Cerny and Davidek (2003) investigated the efficiency of this precursor system inthe generation of the meat-like odourants 2-methyl-3-furanthiol (MFT) and 3-mercapto-2-pentanone (MP) relative to their formation from ribose/cysteine. Reacting 13C5-labelledribose together with unlabelled norfuraneol and cysteine under aqueous conditions (pH 5,95◦C for 4 hours), the authors showed that mainly 13C5-labelled MFT was formed, suggest-ing that norfuraneol is less important as intermediate of MFT. In contrast, MP was foundunlabelled and hence originated from norfuraneol. On the basis of these results, as well asadditional mechanistic studies using the CAMOLA technique (heating of ribose and [13C5]-ribose (1 + 1) with cysteine under above-mentioned conditions), a new reaction pathway forthe formation of MFT and MP from ribose via 1,4-dideoxyosone was proposed (Fig. 3.4).



Ribose

O

OOH O

OH

OH

O

OHSH

OH

O

SH

OH

O

SH

O

OH SH

+ Cys

N

OH SH

SHO

HO

- CO2

NH2

SH SH

NH O

SH

SH

O

Red.

1,4-Dideoxyosone

2-Methyl-3-furanthiol

3-Mercapto-2-pentanone

OOH

O

O

OH

O

O

O

OH

SH

+ H2S

OH SH

O

OH

- H2O

+ H2S

- H2O

+ H2O

- H2O- H2O

- H2O

- H2O

SH

N

SH

-

SH

O

- NH3

+ H2O

+ H2S

Fig. 3.4 Proposed formation pathway for 3-mercapto-2-pentanone and 2-methyl-3-furanthiol from riboseand cysteine via the 1,4-dideoxyosone route (adapted from Cerny and Davidek, 2003).



3.2.2 Carbohydrate fragmentation

Carbohydrate fragments have been found to originate from deoxyosones, ARPs or HRPs,as well as from the sugar directly (Ledl and Schleicher, 1990; Weenen, 1998) (Fig. 3.1).The extent of the sugar cleavage reactions depends on the pH and on the reaction medium,with fragmentation favoured at higher pH values (pH ≥7) and in aqueous systems. Usingisotopic labelling of the sugar molecule in conjunction with GC-MS analysis, C5/C1, C4/C2

and C3/C3 fission reactions were established (review by Tressl et al., 1995). The proposedcleavage routes involve retro-aldolisation, vinylogous retro-aldolisation, �- and �-dicarbonylcleavage (reviewed by Weenen, 1998; Tressl and Rewicki, 1999). Retro-aldolisation is byfar the most accepted fragmentation route.

Studying the generation of acetic acid (which is a major carbohydrate fragmentationproduct and also a good marker for the 2,3-enolisation pathway, as exclusively formed from1-deoxy-2,3-diuloses) during Maillard reaction of glucose and glycine under aqueous con-ditions (90–120◦C, pH 6–8), Davidek et al. (2006a) revealed that the organic acid is mainlyformed through a hydrolytic �-dicarbonyl cleavage pathway (Fig. 3.5). The authors alsoevidenced that �-dicarbonyl cleavage, which can be seen as an acyloin cleavage or a reverseClaisen-type reaction, represents a general cleavage route for diacylcarbinol intermediates ofthe Maillard reaction under aqueous conditions and that the frequently reported hydrolytic�-dicarbonyl cleavage can be ruled out as a sugar fragmentation mechanism. Furthermore,a new sugar fragmentation pathway has been suggested to occur under oxidative conditions(Davidek et al., 2006b) by oxidative �-dicarbonyl cleavage with a Baeyer–Villiger-type rear-rangement as key steps. The sugar degradation pathway will mainly depend on the reactionconditions (Fig. 3.6).

Carbohydrate cleavage products have been analysed by several authors (Nedvidek et al.,1992; Weenen and Apeldoorn, 1996; Hofmann, 1999). Since these intermediates are reactivedicarbonyl and hydroxycarbonyl compounds, trapping agents such as 1,2-diaminobenzeneand ethoxamine hydrochloride were used to transform them into stable quinoxaline andO-ethyloxime derivatives, respectively. Weenen and Apeldoorn (1996) studied the forma-tion of glyoxal, methylglyoxal, 2,3-butanedione and 2,3-pentanedione in both Maillard andcaramelisation reactions. The results are shown in Table 3.1. The study revealed that sugarfragmentation is highest in the presence of a Strecker-inactive amine functionality (cyclo-hexylamine), followed by a Strecker-active amino acid (alanine). Without amine (carameli-sation reaction), the extent of fragmentation was even lower and no detectable amounts of2,3-butanedione and 2,3-pentanedione were observed. The yields of the pentanedione wererelatively high in alanine model systems, indicating that the Strecker aldehyde, acetaldehyde,is involved in its formation. In addition, the ARP of glucose and alanine was found to bean efficient �-dicarbonyl precursor, whereas the 3-deoxyglucosone/alanine reaction mixtureyielded only low concentrations of fission products (Table 3.1). The latter result is in goodagreement with the finding that 3-deoxyglucosone is also a poor pyrazine precursor (Weenenand Tjan, 1994).

Hofmann (1999) studied the time course of the formation of carbohydrate degradationproducts in thermally treated solutions of either xylose or glucose with alanine. The authorshowed that during the first 10 minutes of the Maillard reaction, glyoxal is the most abundantfragmentation product from both xylose and glucose. Its formation can be explained by retro-aldol cleavage of 2-xylosulose or 2-glucosulose. After 20 minutes of the Maillard reaction,the main fission products were found to be methyl glyoxal and hydroxy-2-propanone, withxylose yielding higher amounts than glucose. One year later, Yaylayan and Keyhani (2000)



1-D

eoxy

-2,3

-hex

odiu

lose

CO

CH

3

HH

CC

O OH

OH

CH

2O

H

C

CH

3CO

OH

1-D

eoxy

-2,4

-hex

odiu

lose

HO

H

CC

HO

H

OH

CH

2O

H

C

+

H2O

Ace

tylfo

rmoi

n

H

CC

HO

H

OH

CH

2O

H

CH

O

Ery

thro

se

CC

OO

CH

2OH

CH

3

CC

OH

O

CH

3

H

CH

O

H

CC

HO

H

OH

+

1-D

eoxy

-2,3

-tet

rodi

ulos

e2-

Hyd

roxy

-3-o

xobu

tana

l

OH

H+

+H

OH

CC

OOH

CH

3

C

H

CC

OOH

CH

3

CH

O

H

CC

HO

H

OH

+C

H3C

OO

H

OH

H+

CH

3CO

OH

+OH

OH

H

CCH

3

C

OC

CC

OH

O

CH

3

H

CH

O

OH C

H3C

OO

H

+H

CC

HO

H

OH

2-H

ydro

xy-3

-oxo

buta

nal

2-H

ydro

xy-3

-oxo

buta

nal

1 6

OH

HCC

H3

CC

O OO

CH

3

CO

H

H+

1 6

OH

H

OCC

H3

CC

O O

CH

3

C

OH

H+

1 6

63

63

63

3

6

63

6

3

11

1

4

4

4

6

OH

HCC

H3

HCC

O OH

O

CH

2O

H

CO

H

H+

61

BC

A

CH

3CO

OH

1

Ace

tylfo

rmoi

n

CH

2OH

CH

O

6

CH

2OH

CH

O3

CH

2OH

CH

O

4

CH

3CO

OH

1

CC=

O

HO

H

CH

2O

H

CH

2O

H

Tet

rulo

se63

Fig.3

.5Fo

rmat

ion

ofac

etic

acid

from

gluc

ose

via

1-de

oxy-

2,3-

hexo

diul

ose

asth

eke

yin

term

edia

teby

the

hydr

olyt

ic�

-dic

arbo

nylc

leav

age

mec

hani

sm,i

ndic

atin

gth

eva

rious

poss

ible

degr

adat

ion

path

way

s(a

dapt

edfro

mD

avid

eket

al.,

2006

a).



1-Deoxy-2,3-hexodiulose


C O

CH3

H

H

C

C

O

OH

OH

CH2OH

C

C

C

H OH

O

CH2OH

CH2OH

Tetrulose

B1

OHH C

CH3

H C

C

O

OH

O

CH2OH

C

+

Acetic acid

CH3

O

CH2OH

C

Acetol

A1C

C

H OH

OH

CH2OH

COOH

H

CH3

COOH

+

Erythronic acid

CH OH

CH2OH

COOH

Glyceric acid

B2 +

A2

CH3

OHC

OC

H C

C

OH

O

CH2OH

H

+

Lactic acid

CH3

OHC

COOH

H


Fig. 3.6 Scheme summarising possible pathways of sugar fragmentation via 1-deoxyhexodiuloses. A1and A2, oxidative �-dicarbonyl cleavage; B1 and B2, hydrolytic �-dicarbonyl cleavage by nucleophilicattack of OH− at the C-2 and C-4 carbonyl group, respectively (adapted from Davidek et al., 2006b).

performed a mechanistic study to determine the origin of Maillard intermediates such asglycolaldehyde, methylglyoxal, hydroxy-2-propanone and 3-hydroxy-2-butanone.

2,3-Butanedione (diacetyl) and 2,3-pentanedione are aroma-active carbohydrate cleav-age products, contributing to a sweet-caramel odour of coffee (Grosch, 2001), and, forexample, in the presence of hydrogen sulfide or cysteine, they are precursors of importantsulfur aroma compounds such as 2-mercapto-3-butanone and 2/3-mercapto-3/2-pentanone(Hofmann, 1995). Yaylayan and Keyhani (1999) investigated the origin of these two dicar-bonyl compounds in glucose/alanine Maillard model systems under pyrolytic conditions.Using labelled glucose or alanine, the authors showed that 90% of the formed pentane-dione requires the participation of the C2/C3 atoms of alanine, whereas diacetyl was derivedfrom the sugar chain only. This result is supported by the finding of Hofmann (1995) that



Table 3.1 Formation of �-dicarbonyl products (according to Weenen and Apeldoorn, 1996).

α-Dicarbonyl products (µg)

Amine Carbohydrate GlyoxalMethylglyoxal

2,3-Buta-nedione

2,3-Penta-nedione

No Glucose 26 11 – –No Fructose 28 15 – –No Xylose 62 17 – –No 3-Deoxyglucosone 23 57 – –No Fru-Alaa 103 101 98 18Alanine Glucose 58 43 41 38Alanine Fructose 45 28 22 25Alanine Xylose 27 81 28 42Alanine 3-Deoxyglucosone 16 56 11 21Alanine Fru-Alaa 81 67 81 22Cyclohexylamine Glucose 618 865 227 39Cyclohexylamine Fructose 691 1104 265 89Cyclohexylamine Xylose 591 925 614 101Cyclohexylamine 3-Deoxyglucosone 317 583 146 25Cyclohexylamine Fru-Alaa 509 454 232 39

1 N-1-(deoxy-d-fructosyl)-l-alanine (Amadori rearrangement product of glucose and alanine).

2,3-pentanedione is formed by reacting acetaldehyde and hydroxy-2-propanone. Schieberleet al. (2003) applied the CAMOLA technique to a Maillard model system of glucose (13C6 +12C6 = 1 + 1) and proline and showed that 2,3-butanedione was formed by 87 and 13%through the recombination of C3 + C1 and C2 + C2 fragments (Route B in Fig. 3.1), respec-tively. As a result, no diacetyl is generated from the intact carbon chain (Route A in Fig. 3.1).

3.2.3 Strecker degradation

In the presence of �- or vinylogous dicarbonyl compounds, �-amino acids can undergoa ‘decarboxylating transamination’, which results in the formation of aldehydes with onecarbon atom less than the amino acid (Strecker aldehydes) (Schonberg and Moubacher,1952). The Strecker degradation of amino acids is a key reaction in the generation of potentaroma compounds during Maillard-type processes (Ledl and Schleicher, 1990) (see alsoFig. 3.1). Certain amino acids (leucine, valine, methionine or phenylalanine) are knownto produce Strecker aldehydes with significant odour strength such as 3-methylbutanal,methylpropanal, methional or phenylacetaldehyde. These aldehydes have been confirmed askey contributors to many thermally processed foods (Hofmann et al., 2000). Besides aldehydeformation, Strecker degradation also contributes to flavour formation during Maillard reactionby reducing dicarbonyls to hydroxycarbonyls (e.g. formation of 1,4-dideoxyosone from 1-deoxyosone) (Nedvidek et al., 1992) or by generating �-aminocarbonyl compounds, whichare pyrazine precursors (Weenen and Tjan, 1994).

Weenen and van der Ven (1999) studied the formation of phenylacetaldehyde in Maillardmodel systems, including reactions of phenylalanine with various sugars, �-dicarbonyl andhydroxycarbonyl compounds as well as ARPs. The authors found that methyl glyoxal wasthe most efficient dicarbonyl compound for the formation of phenylacetaldehyde, followedby 3-deoxyerythrosone, glyoxal, 3-deoxyxylosone and 3-deoxyglucosone. Hydroxycarbonylcompounds such as dihydroxyacetone and glyceraldehyde also yielded high amounts of the



COOH O2,Me2+

H+O

HO

HO

OHOH

NH

COOH

O

O

O

OH

OH

HOH

HO

HO

OHOH

N

COOH

O

HO

HO

OHOH

N

Me

O

O

+H

Me2+,H2O2

CO2,H2O

H+

H2O

NH2

O

OH

OH

HO

NO

OH

OH

OH

HO

N

COOH

Fig. 3.7 Formation of phenylacetaldehyde via an oxidative degradation of N-(1-deoxy-D-fructosyl)-L-phenylalanine (Fru-Phe) (adapted from Hofmann and Schieberle, 2000a).

Strecker aldehyde. Sugars were less reactive, with the reactivity decreasing in the order ery-throse, xylose, fructose and glucose. In addition, the authors showed that the Amadori com-pound Fru-Phe is a superior phenylacetaldehyde precursor to the corresponding sugar/aminoacid mixture. This finding was also confirmed by Hofmann and Schieberle (2000a).

In addition, Hofmann and Schieberle (2000a) revealed that the yields of phenylacetalde-hyde formed from the Amadori compound Fru-Phe were significantly increased in thepresence of oxygen and copper (II) ions. On the basis of the observation that 1,2-hexodiuloseis also generated in high amounts under these conditions, they proposed a mechanism forthe formation of phenylacetaldehyde from Fru-Phe (Fig. 3.7). Hofmann et al. (2000) addi-tionally found that in the reaction of phenylalanine and glucose, considerable amounts ofphenylacetic acid were generated. While the formation of phenylacetaldehyde showed anoptimum pH of 5 and was not influenced by oxygen, the acid was most abundant at pH 9in the presence of oxygen and copper (II) ions. The authors proposed a mechanism for thegeneration of phenylacetic acid that involves a similar oxidation step to that shown in Fig. 3.7.

Cremer and Eichner (2000) studied the influence of the pH on the formation of3-methylbutanal during Maillard reaction of glucose and leucine. They showed that theformation rate of the aldehyde was higher at pH 7 than at pH 5 or 3. This was consistentwith the high degradation rate of the Amadori compound Fru-Leu at pH 7, suggesting thatthe ARP was a good precursor for 3-methylbutanal. However, Chan and Reineccius (1994)showed that the optimal reaction conditions of different Strecker aldehydes vary, owing todifferences in stability of the aldehydes.

3.2.4 Interactions with lipids

In addition to the Maillard reaction, lipid oxidation is another major reaction occurringin process flavour production and food systems. Both reaction cascades include a wholenetwork of different reactions in which extraordinary complex mixtures of compounds areobtained, triggering important changes in food flavour, colour, texture and nutritional value,with desirable and undesirable consequences. In addition, both reactions are intimatelyinterrelated as shown in Fig. 3.8 (Zamora and Hidalgo, 2005) and the products of eachreaction influence the other. Furthermore, there are common intermediates and products inboth pathways. The existing data suggest that both Maillard reaction and lipid peroxidationare so closely interrelated that they should be considered simultaneously to understand the



Aroma Taste Antioxidants Coloured compounds Neo-formed toxicants

Lipid

Oxidation

Lipid hydroperoxides

Reducingsugar

Oxidation

Aminocompound

Amadori rearrangement

1-Amino-1deoxy-2-ketose

Sugar dehydrationSugar fragmentation

Sugar enolisationStrecker degradation

GlyoxalMethylglyoxal

Others Volatile and non-volatile monomers

• Schiff base of HMF or furfural• Reductones• Fission products• Aldehydes• Others

N-Substitutedglycosylamine

Lipid peroxyl and hydroxyl radicals

Formation of stable compoundsPolymerisation

Cleavage

Cleavage

Aldol condensationCarbonyl-amine polymerisation

Pyrrole polymerisation

Carbonyl-amine polymerisation Aldol condensation

Carbonyl-amine polymerisation

KetonesAlcoholsEpoxidesDimers

Polymers

AldehydesKetonesAlcoholsEpoxides

HydrocarbonsAcids

Hydroxyalkylpyrroles

Cyclic peroxidesHydroperoxycompounds

Hydroperoxide decomposition

Fig. 3.8 Known interactions between Maillard reaction and lipid oxidation pathways in nonenzymaticbrowning development (adapted from Zamora and Hidalgo, 2005).

reaction mechanisms, kinetics, and products in the complex mixtures of carbohydrates, lipidsand proteins occurring in food systems and process flavours. In these systems, lipids andcarbohydrates are competing in the chemical modification of amino compounds (e.g. proteinsand phospholipids). Therefore, although there are significant differences between Maillardreaction and lipid peroxidation, many aspects of both reactions can be better understood ifthey are included in only one general carbonyl pathway that can be initiated by both lipidsand carbohydrates (Hidalgo and Zamora, 2005).

As an example, 2-pentylpyridine has been reported (Henderson and Nawar, 1981) as aninteraction product of linoleic acid and valine, with 2,4-decadienal and ammonia being thekey intermediates (Fig. 3.9). Its formation was studied by Kim et al. (1996) in model systems

R

NH2

COOH

COOH

Amino acid (asparagine, glutamine)

Linoleic acid

O

Maillard

reaction

Lipid

Oxidation

NH3

Ammonia

(E,E)-2,4-Decadienal

N

NH[O]

NH

Fig. 3.9 Formation of 2-pentylpyridine from 2,4-decadienal and an amino acid (adapted fromHenderson and Nawar, 1981).



by reacting 2,4-decadienal with amino acids (glycine, aspartic acid, asparagine, glutamicacid and glutamine) at 180◦C for 1 hour (pH 7.5). The relative yields of alkylpyridineformation from the reactions were asparagine � glutamine � aspartic acid � glutamicacid � glycine. When amide-15N-labelled glutamine and asparagine were heated with 2,4-decadienal, the relative contribution of amide nitrogens to the formation of alkylpyridinewas determined. Approximately half the nitrogen atoms in 2-pentylpyridine formed fromasparagine, originated from the amide nitrogens of asparagine, whereas when glutamine wasthe reactant, almost all the nitrogen atoms came from the amide nitrogens in glutamine. Theabove-mentioned results may indicate that both free ammonia and �-amino groups boundin amino acids can contribute to the formation of alkylpyridines, but free ammonia does somore effectively.

The formation of Strecker aldehydes in the presence of lipid oxidation products isanother example, illustrating the interactions between Maillard and lipid intermediates.Strecker degradation of amino acids is one of the most important reactions leading to fi-nal aroma compounds in the Maillard reaction. Hidalgo and Zamora (2004) have studiedthe reaction of 4,5-epoxy-2-alkenals with phenylalanine. In addition to N-substituted 2-(1-hydroxyalkyl)pyrroles and N-substituted pyrroles, which are major products of the reac-tion, the formation of both the Strecker aldehyde, phenylacetaldehyde, and 2-alkylpyridineswas also observed. The aldehyde was only produced from the free amino acid (not esteri-fied), suggested to be produced through imine formation, which is then decarboxylated andhydrolysed (Fig. 3.10). This reaction also produces a hydroxyl amino derivative, which is theorigin of the 2-alkylpyridines. These data indicate that Strecker-type degradation of aminoacids occurs at low temperature by some lipid oxidation products. This is a proof of theinterrelations between lipid oxidation and Maillard reaction, which are able to produce com-mon products by analogous mechanisms. However, recent results suggest that, analogouslyto carbohydrates, certain lipid oxidation products may also degrade certain amino acids to

H

+

+

H2N− H2O

+ H2O

O

H

− CO2

OH

R1

R1 N

OO

N

H

OOH

OO

O

5 6

14

13

15

OH

R1

R1

R1

NH2

N

Fig. 3.10 Strecker-type degradation of phenylalanine produced by 4,5-epoxy-2-alkenals (adapted fromHidalgo and Zamora, 2004).



undesirable compounds, e.g. vinylogous derivatives such as styrene (Hidalgo and Zamora,2007).

Simulating food flavours by the process flavour approach requires precursors and recipesthat are close to food or which well represent the composition of food. Therefore, lipidsare key ingredients in the reaction flavour system to obtain boiled chicken notes. Similarly,polyphenols play an important role in generating cocoa flavour. The role of these specificcomponents (lipids, polyphenols, vitamins) may be (i) generating new specific flavour com-pounds (Whitfield, 1992) as compared to the pure Maillard system (e.g. 2-pentylpyridine)or (ii) intervening in the Maillard reaction cascade and thus alerting the overall flavourcomposition. The latter is probably of higher relevance.

3.3 IMPORTANT AROMA COMPOUNDS DERIVEDFROM MAILLARD REACTION IN FOOD ANDPROCESS FLAVOURS

In the flavour industry, process flavours are often developed by empirical means, i.e. thereaction conditions are optimised using organoleptic evaluation. However, this approachshould be accompanied by analytical evaluation of the products. Knowledge of the impor-tant aroma compounds derived from the Maillard reaction in food and process flavours isessential in order to focus investigations into reaction mechanisms enhancing the key aromacompounds. In addition, the knowledge of their formation pathways and key intermediatesallows the development of multi-step approaches (e.g. two-step reactions), which aim at op-timising reaction conditions of each of the flavour generation stages (e.g. from sugar/aminoacid mixture to deoxyosones, from deoxyosones to target flavour compounds). One of thesesteps (often the first step) can also be a biotransformation that yields an intermediate that,on thermal treatment, releases the key aroma compound. An example for such an approachis the biogeneration of 2-(1-hydroxyethyl)-4,5-dihydrothiazole through yeast fermentation,which releases 2-acetylthiazoline on microwave heating (Bel Rhild et al., 2002).

It is recommended that the evaluation of character-impact aroma compounds should bebased on a combination of instrumental and sensorial analysis. A well-established approachinvolves the determination of the odour activity values (ratio of concentration and odourthreshold values) of aroma compounds (Grosch, 1994). This requires quantitative analysis ofa selected number of aroma compounds, the importance of which has been screened by GC-olfactometry (GC-O). Suitable screening techniques such as aroma extract dilution analysisor headspace dilution analysis are available (Ullrich and Grosch, 1987; Guth and Grosch,1993a, b). As these methods do not consider interactions of different aroma compounds, theyshould be combined with organoleptic evaluations of reconstituted model mixtures.

Our selection of important Maillard-derived aroma compounds (see Sections 3.3.1 and3.3.2) is primarily based on results of GC-O techniques or other sensory evaluations or onquantitative data.

3.3.1 Character-impact compounds of thermallytreated foods

Character-impact compounds of thermally treated foods that are formed during Maillardreaction are summarised in Table 3.2. Their identification in foodstuffs such as meat, bread,


66 Food Flavour TechnologyTa

ble

3.2

Impo

rtant

arom

aco

mpo

unds

deriv

edfro

mth

eM

ailla

rdre

actio

nin

vario

usth

erm

ally

treat

edfo

ods.

a

Odour

Odour

thre

shold

inN

o.

Com

pound

des

crip

tion

wate

r(�

g/k

g)b

Det

ecte

din

b

A:C

ompo

unds

cont

aini

ngsu

lfur

12-

Met

hyl-3

-fura

nthi

olM

eaty

,sw

eet,

sulfu

ry0.

007

(1)

Mea

t(2–6

),ye

ast(

20,

40

),co

ffee

(7,

8)

22-

Furfu

rylth

iol

Roas

ty,s

ulfu

ry0.

01(1

)M

eat(

3–6

,9

),ye

ast(

40

),co

ffee

(7,

8),

sesa

me

(10

),po

pcor

n(1

1)

33-

Mer

capt

o-2-

buta

none

Sulfu

ry,c

atty

3.0

(1)

Beef

(12

)

43-

Mer

capt

o-2-

pent

anon

eSu

lfury

,cat

ty0.

7(1

)Be

ef(3

,4,

13

),ch

icke

n(3

,14

),ye

ast(

40

)

52,

5-D

imet

hyl-3

-fura

nthi

olM

eaty

,sw

eet,

sulfu

ry0.

018

(1)

Chi

cken

(4)

62-

Met

hyl-3

-(met

hylth

io)fu

ran

Mea

ty,t

hiam

ine-

like

0.05

(15

)Ye

ast(

15

)

72-

Met

hyl-3

-(met

hyld

ithio

)fura

nC

ooke

dm

eat-l

ike

0.00

4(1

6)

Coc

oa(1

7),

choc

olat

e(1

7),

mea

t(41

)

82-

Met

hyl-3

-(met

hyltr

ithio

)fura

nC

ooke

dm

eat-l

ike

—Be

ef(1

2)

92-

Furfu

rylm

ethy

ldis

ulfid

eRo

asty

,bro

thy

—Be

ef(1

2)

102-

Met

hyl-3

-fury

lthio

acet

ate

Mea

ty,o

nion

-like

,cof

fee-

like

—Ye

ast(

20

)

111-

(2-M

ethy

l-3-fu

rylth

io)-e

than

ethi

olM

eaty

,roa

stbe

ef—

Yeas

t(20

)

124-

Hyd

roxy

-2,5

-dim

ethy

l-3(2

H)-

thio

phen

one

Car

amel

-like

,fru

ity0.

05(2

0)

Yeas

t(20

)

13M

ethi

onal

Coo

ked

pota

to-li

ke0.

2(1

8)

Beef

(3,

9,

19

),ch

icke

n(4

,5

),po

rk(6

),co

ffee

(7),

Fren

chfri

es(2

1),

pota

toch

ip(1

8),

fish

(22,

23

),br

ead

(24

),ye

ast(

40

)

142-

Ace

tyl-2

-thia

zolin

eRo

asty

,pop

corn

-like

,bur

nt1.

0(2

5)

Beef

(3,

9),

chic

ken

(4,

5),

sesa

me

(10

),fis

h(2

2)

B:C

ompo

unds

cont

aini

ngox

ygen

154-

Hyd

roxy

-2,5

-dim

ethy

l-3(2

H)-f

uran

one

Car

amel

-like

,stra

wbe

rry-

like

10(2

6)

Beef

(3,

9,

19

),ch

icke

n(6

),co

ffee

(7),

beer

(27

),po

pcor

n(2

7),

brea

d(2

4),

choc

olat

e(1

7),

sesa

me

(28

),Fr

ench

fries

(21

),te

a(2

9),

yeas

t(40

)

163-

Hyd

roxy

-4,5

-dim

ethy

l-2(5

H)-f

uran

one

(sot

olon

)Se

ason

ing-

like

0.3

(5)

Beef

(3,

9,

19

),ch

icke

n(5

),co

ffee

(7),

Fren

chfri

es(2

1),

tea

(29

),ch

ocol

ate

(17

),co

coa

(17

),ye

ast(

40

)17

2-Et

hyl-4

-hyd

roxy

-5-m

ethy

l-3(2

H)-

fura

none

Car

amel

-like

,sw

eet

1.15

(26

)C

offe

e(7

)

183-

Hyd

roxy

-4-m

ethy

l-5-e

thyl

-2(5

H)-

fura

none

(abh

exon

)Se

ason

ing-

like

7.5

(26

)C

offe

e(2

6),

choc

olat

e(1

7),

coco

a(1

7)



192/

3-M

ethy

lbut

anal

Mal

ty,c

ocoa

-like

0.35

(5)

Mea

t(3,

5),

yeas

t(40

),ch

ocol

ate

(17

),co

coa

(17

),co

ffee

(7),

Fren

chfri

es(2

1),

brea

d(2

4),

tea

(29

),be

er(3

0)

20M

ethy

lpro

pana

lM

alty

,fru

ity,p

unge

nt0.

7(2

3)

Beef

(19

),ch

icke

n(5

),br

ead

(24

),ch

ocol

ate

(17

),co

ffee

(7),

Fren

chfri

es(2

0),

tea

(29

)

21Ph

enyl

acet

alde

hyde

Hon

ey-li

ke,s

wee

t,flo

wer

y4

(18

)Be

ef(2

,19

),co

coa

(17

),ch

ocol

ate

(17

),br

ead

(24

),Fr

ench

fries

(21

),co

ffee

(7),

tea

(29

),ye

ast(

40

)

22A

ceta

ldeh

yde

Solv

ent-l

ike

25(9

)Be

ef(9

),ch

icke

n(5

),co

ffee

(8),

Fren

chfri

es(2

1),

fish

(31

)

232,

3-Bu

tane

dion

eBu

ttery

15(3

2)

Beef

(3,

9),

brea

d(2

4),

coffe

e(7

),fis

h(2

2,

23

),C

hoco

late

(17

),co

coa

(17

),ye

ast(

40

)

242,

3-Pe

ntan

edio

neBu

ttery

,gre

en30

(32

)Fi

sh(2

2,

23

),co

ffee

(7),

brea

d(2

4)

C:C

ompo

unds

cont

aini

ngni

troge

n25

2-A

cety

l-1-p

yrro

line

Roas

ty,p

opco

rn,b

read

-like

0.1

(33

)Br

ead

(24

),ric

e(3

4),

popc

orn

(11

),se

sam

e(1

0),

beef

(2,

3),

Fren

chfri

es(2

1),

yeas

t(40

)

266-

Ace

tylte

trahy

drop

yrid

ine

Roas

ty,c

rack

er-li

ke1.

6(3

5)

Brea

d(2

4),

popc

orn

(11

)

272-

Ethy

l-3,5

-dim

ethy

lpyr

azin

eEa

rthy,

roas

ty0.

16(7

)Be

ef(1

9,

25

),ch

icke

n(1

4),

coffe

e(7

),se

sam

e(1

0),

brea

d(2

4),

Fren

chfri

es(2

1),

coco

a(1

7),

choc

olat

e(1

7),

popc

orn

(36

)

282-

Ethy

l-3,6

-dim

ethy

lpyr

azin

eEa

rthy,

roas

ty0.

4(1

8)

Brea

d(2

4),

Fren

chfri

es(2

1),

coco

a(1

7),

choc

olat

e(1

7),

popc

orn

(36

)

292,

3-D

ieth

yl-5

-met

hylp

yraz

ine

Earth

y,ro

asty

0.09

(7)

Beef

(19,

25

),ch

icke

n(1

4),

coffe

e(7

),Fr

ench

fries

(21

),co

coa

(17

),ch

ocol

ate

(17

),se

sam

e(1

0),

popc

orn

(36

),br

ead

(37

),ye

ast(

40

)

302-

Ethe

nyl-3

,5-d

imet

hylp

yraz

ine

Earth

y,ro

asty

—C

offe

e(7

)

312-

Ethe

nyl-3

-eth

yl-5

-met

hyl-p

yraz

ine

Earth

y,ro

asty

—C

offe

e(7

),Fr

ench

fries

(21

)

322-

Ace

tylp

yraz

ine

Roas

ty,s

wee

t,nu

tty62

(38

)Se

sam

e(2

8),

popc

orn

(39

),br

ead

(36

)

aTh

ese

nsor

ysi

gnifi

canc

eof

the

arom

aco

mpo

unds

was

asse

ssed

byqu

antit

ativ

eda

taor

byG

C-O

tech

niqu

es.

b1

,Hof

man

n(1

995)

;2,G

asse

rand

Gro

sch

(198

8);3

,Ker

sche

rand

Gro

sch

(199

7);4

,Gas

sera

ndG

rosc

h(1

990)

;5,K

erle

rand

Gro

sch

(199

7);6

,Gas

sera

ndG

rosc

h(1

991)

;7,

Gro

sch

(200

1);8

,Sem

mel

roch

and

Gro

sch

(199

5);9

,Gut

han

dG

rosc

h(1

994)

;10

,Sch

iebe

rle(1

993a

);1

1,S

chie

berle

(199

1a);

12

,Mot

tram

and

Mad

ruga

(199

4);1

3,G

uth

and

Gro

sch

(199

3a);

14

,Ker

ler(

1996

);1

5,M

acLe

odan

dA

mes

(198

6);1

6,S

chie

berle

etal

.(20

00);

17

,Sch

nerm

ann

and

Schi

eber

le(1

997)

;18

,Gua

dagn

ieta

l.(1

972)

;19

,Ker

lera

ndG

rosc

h(1

996)

;20

,Wer

khof

feta

l.(1

991)

;21

,Wag

nera

ndG

rosc

h(1

997)

;22

,Milo

and

Gro

sch

(199

3);2

3,M

iloan

dG

rosc

h(1

996)

;24

,Ryc

hlik

and

Gro

sch

(199

6);2

5,C

erny

and

Gro

sch

(199

3);2

6,S

emm

elro

chet

al.(

1995

);2

7,S

chie

berle

(199

3b);

28

,Sch

iebe

rle(1

996)

;29

,Gut

han

dG

rosc

h(1

993b

);3

0,S

chie

berle

(199

1b);

31

,Milo

and

Gro

sch

(199

5);

32

,Bla

nket

al.(

1992

);3

3,B

utte

ryet

al.(

1983

);3

4,B

utte

ryet

al.(

1982

);3

5,B

utte

ryan

dLin

g(1

995)

;36

,Sch

iebe

rlean

dG

rosc

h(1

987)

;37

,Sch

iebe

rlean

dG

rosc

h(1

994)

;38

,Te

rani

shie

tal.

(197

5);3

9,S

chie

berle

(199

5);4

0,M

unch

and

Schi

eber

le(1

998)

;41

,Mad

ruga

and

Mot

tram

(199

5).



coffee, cocoa, chocolate, sesame, popcorn, French fries, tea, fish and yeast, as well as odourdescription and odour thresholds of the aroma substances, is covered. In the followingdiscussion of Table 3.2, emphasis is given to important aroma compounds of meat, yeast,coffee and bread.

The meaty character of boiled beef, pork or chicken is mainly due to sulfur compoundssuch as MFT, 2-furfurylthiol, 3-mercapto-2-butanone, 2/3-mercapto-3/2-pentanone, DMFT,methanethiol, hydrogen sulfide and methional (Gasser and Grosch, 1988, 1990; Mottram andMadruga, 1994; Kerscher, 2000). The precursors of these aroma substances in meat are knownto be free or bound C5-sugars such as ribose, ribose phosphate and inosine monophosphate aswell as sulfur-containing compounds such as thiamine, cysteine, glutathione and methionine.After the publication of the patent of Morton et al. (1960), in which meat aroma formationestablished from the reaction of cysteine and ribose was described, a great number of patentsand publication on meat-like process flavours followed (see Section 3.4.3).

It is well known that species-specific differences in the aroma of cooked meats suchas beef and chicken are mainly due to concentration and composition differences in lipid-derived flavour substances. Kerscher and Grosch (1998) and Kerscher (2000) confirmedthese findings and showed that significant differences also exist for Maillard-derived aromacompounds. Cooked beef, for example, was found to contain higher amounts of the sulfurcompounds MFT and 2-furfurylthiol as well as the caramel-like 4-hydroxy-2,5-dimethyl-3(2H)-furanone, whereas MP and methional were more important in cooked chicken.

The character-impact compounds of yeast extracts were found to be very similar to thoseof cooked meat, which is due to similar pools of Maillard precursors. Munch and Schieberle(1998) reported high odour activity values for the sulfur compounds MFT, 2-furfurylthiol,MP and methional, the Strecker aldehydes 3-methylbutanal and phenylacetaldehyde, as wellas the furanones 4-hydroxy-2,5-dimethyl-3(2H)-furanone (Furaneol) and 3-hydroxy-4,5-dimethyl-2(5H)-furanone (sotolon). Werkhoff et al. (1991) additionally identified 2-methyl-3-(methylthio)furan, 2-methyl-3-furylthioacetate, 1-(2-methyl-3-furylthio)ethanethiol and 4-hydroxy-2,5-dimethyl-3(2H)-thiophenone (thiofuraneol) in yeast. The authors claimed thatthese sulfur compounds contribute significantly to the meaty character of yeast. Besidesyeast, onion extracts are also used in process flavourings. Widder et al. (2000) have iden-tified 3-mercapto-2-methylpentan-1-ol as new powerful aroma compounds in both processflavourings (containing onion extract) and raw onions. This odourant that exhibits a pleasantmeat broth, sweetish, onion and leek-like character (at a low concentration of 0.5 ppb in wa-ter), is suggested to be formed by aldol condensation of two molecules of propanal, followedby hydrogen sulfide addition at the double bond to yield 3-mercapto-2-methylpentanal. Thealdehydes are finally reduced enzymatically to the corresponding alcohol.

The Maillard reaction is also a key reaction in flavour formation during roasting ofcoffee. The precursor pool in green coffee comprises a complex mixture of various solublesugars such as glucose, fructose, galactose and sucrose. In addition, the amount of polymericarabinose and rhamnose was found to decrease during roasting, which indicates that thesesugars are also involved in caramelisation and Maillard processes (Tressl, 1989). The totalamino acid content drops by about 30% during roasting. Especially, the amino acids –lysine, serine, threonine, arginine, histidine, methionine and cystine – are degraded to ahigh extent during the roasting process (Belitz et al., 2008). Semmelroch and Grosch (1995)reported the simulation of the aroma of Arabica and Robusta coffee brews using reconstitutedmixtures of 23 aroma compounds. The authors showed that the Maillard products – 2-furfurylthiol, Furaneol, sotolon, methanethiol, 2,3-butanedione, 2,3-pentanedione, 2-ethyl-3,5-dimethylpyrazine (EDMP), 2,3-diethyl-5-methylpyrazine (DEMP), methylpropanal and



O

N+

O

OH

O

OH

OO

O

N

O

N

OH

OH

OH

ONH

COOH

COOH

OH

N+

H CO

OH

OH

O

5-MPC

5-MPF

Alapyridaine

Dru-Glu1-Oxo-2,3-dihydro-1H-

indolizinium-6-olate

O

N+

O

OH

O

OH

OO

O

N

O

N

OH

OH

OH

ONH

COOH

COOH

OH

N+

H CO

OH

OH

O

5-MPC

5-MPF

Alapyridaine

Dru-Glu1-Oxo-2,3-dihydro-1H-

indolizinium-6-olate

Fig. 3.11 Chemical structures of some recently identified taste molecules formed by Maillard-type reac-tions. Fru-Glu, N-(1-deoxy-D-fructos-1-yl)-L-glutamic acid; 5-MPC, 5-methyl-2-(1-pyrrolidinyl)-2-cyclopenten-1-one; MPF, and 5-methyl-4-(1-pyrrolidinyl)-3(2H)-furanone.

3-methylbutanal – contribute to the flavour of coffee brews. They also investigated the aromadifferences between Arabica and Robusta coffee. The more earthy/roasty and less caramelcharacter of Robusta was found to be due to higher concentrations of the pyrazines EDPMand DEMP as well as to lower amounts of Furaneol and sotolon, respectively.

The flavour of cereal products, especially of bread, has been studied extensively, and theresults have been reviewed (Grosch and Schieberle, 1997). 2-Acetyl-1-pyrroline (ACP) and6-acetyltetrahydropyridine (ACTP) are responsible for the pleasant roasty character of wheatbread crust and popcorn. However, these compounds are not important in wheat bread crumband rye bread. Both ACP and ACTP are generated by a reaction of proline with reducingsugars or sugar breakdown products (Schieberle, 1990). When ornithine instead of prolinewas reacted, only ACP was formed. The fact that yeast contains relatively high amounts ofornithine explains why ACP concentrations in bread are strongly dependent on the amount ofyeast used in the baking process (Grosch and Schieberle, 1997). Other important Maillard-derived aroma compounds of wheat and rye bread are Furaneol and the Strecker aldehydesmethional, 3-methylbutanal and methylpropanal. They contribute to the caramel-like andmalty aroma of bread.

Another breakthrough has been the identification of new taste-active molecules and tastemodifiers as result of Maillard-type reactions (Fig. 3.11), which contribute to the over-all flavour perception (Hofmann, 2005). As an example, Alapyridaine has been identifiedin Maillard model systems containing hexose sugars and alanine as well as in beef brothas an essential compound enhancing the sweet taste and umami character in beef broth(Ottinger and Hofmann, 2003; Soldo et al., 2003). Glycoconjugates of glutamic acid, namelythe N-glycoside dipotassium N-(d-glucos-1-yl)-l-glutamate and the corresponding Amadoricompound N-(1-deoxy-d-fructos-1-yl)-l-glutamic acid (Fru-Glu), were found by systematicsensory studies to exhibit a pronounced umami-like taste, with recognition taste thresholds of1–2 mmol/L, close to that of monosodium glutamate (MSG) (Beksan et al., 2003). Contraryto MSG, they do not show the sweetish and slightly soapy by-note, but evoke an intenseumami, seasoning, and bouillon-like taste. Added to a bouillon base, which did not containany taste enhancers, both glycoconjugates imparted a distinct umami character similar to thecontrol sample containing the same amount of MSG on a molar basis (Schlichtherle-Cernyet al., 2002). The Amadori compound Fru-Glu has been reported in dried tomatoes as an ex-ample (Eichner et al., 1994). Furthermore, thermal treatment of aqueous solutions of xylose,



rhamnose and l-alanine led to a rapid development of a bitter taste of the reaction mix-ture (Frank et al., 2003). Liquid chromatography/mass spectrometry (LC/MS) and nuclearmagnetic resonance (NMR) spectroscopy revealed 1-oxo-2,3-dihydro-1H-indolizinium-6-olates as the key compounds and most significant contributors to the intense bitter tasteof this process flavour mixture. Thermally treated glucose/L-proline mixtures that con-tain ‘cooling’ compounds were recently reported. These Maillard systems generate 5-methyl-2-(1-pyrrolidinyl)-2-cyclopenten-1-one (5-MPC) and 5-methyl-4-(1-pyrrolidinyl)-3(2H)-furanone (MPF) as key compounds contributing to the cooling sensation withoutimparting aroma notes (Hofmann et al., 2001; Ottinger et al., 2001a). They have also beenfound in dark malt (10–100 �g/kg) (Ottinger et al., 2001b).

3.3.2 Character-impact compounds of process flavours

Meat-like process flavours are often prepared by reacting cysteine and/or thiamine withsugars, although pentoses such as xylose and ribose are preferably used. There is a rangeof additional precursor sources such as pectin hydrolysates (C-source), hydrolysed veg-etable proteins and wheat protein hydrolysates (N-sources) as well as hydrogen sulfide,inorganic sulfides, onion, garlic and cabbage (S-sources). The products serve as buildingblocks in the creation of meat flavours (see also Section 3.4.3). They can also be describedas middle notes that are combined with base notes (mainly taste compounds and tasteenhancers) and top notes (mainly compounded flavourings) (savoury flavour pyramid ac-cording to Yeretzian et al., 2007). The important sulfur-containing compounds in processflavourings derived from either cysteine/ribose or thiamine reaction systems are quite similar(Table 3.3). The aroma of both cysteine- and thiamine-based process flavours is determinedby MFT, 3-mercapto-2-butanone, 2/3-mercapto-3/2-pentanone, 2-methyl-3-thiophenethioland 2-thenylthiol (Guntert et al., 1992, 1996; Hofmann and Schieberle, 1995, 1997). Thesame compounds are also responsible for the meaty character of process flavours that arebased on 5′-inosine monophosphate (5′-IMP) and cysteine (Zhang and Ho, 1991; Madrugaand Mottram, 1998). Although 5′-IMP is more abundant than ribose in raw meat, it is amuch poorer precursor for these sulfur compounds than ribose, when reacted with cysteine(Mottram and Nobrega, 1998). Other character-impact compounds that are formed primarilyin thiamine or thiamine/cysteine reaction systems are 2-methyl-4,5-dihydro-3-furanthiol, 1-(methylthio)ethanethiol, mercaptoacetaldehyde and 2-methyl-1,3-dithiolane (Guntert et al.,1996).

Many of the thiols mentioned above are also important aroma substances in glu-cose/cysteine or rhamnose/cysteine process flavourings (Hofmann and Schieberle, 1997).However, both reaction systems contain other characteristic aroma compounds. For example,2-(1-mercaptoethyl)furan and its thiophene derivative are only formed from glucose and cys-teine, whereas 4-hydroxy-2,5-dimethyl-3(2H)-furanone (Furaneol) and 3-hydroxy-6-methyl-2(2H)-pyranone belong to key odourants of the rhamnose system. The latter two substancesare responsible for the strong caramel and seasoning-like character of rhamnose/cysteineprocess blocks, which render them very suitable for application in beef flavours. Anothercompound having seasoning-like character is 3-hydroxy-4,5-dimethyl-2(5H)-furanone (so-tolon). Sotolon was found to contribute to the aroma of various cysteine-derived reactionflavours (Hofmann and Schieberle, 1995, 1997). In contrast to the other compounds men-tioned above, the formation of sotolon is less influenced by the type of sugar.



Table

3.3

Cha

ract

er-im

pact

com

poun

dsof

proc

ess

flavo

urin

gs.a

Odour

thre

shold

No.

Com

pound

Odour

des

crip

tion

(�g/k

g)

inw

ate

rbD

etec

ted

inb

A:C

ompo

unds

cont

aini

ngsu

lfur

12-

Met

hyl-3

-fura

nthi

olM

eaty

,sul

fury

,sw

eet

0.00

7(1

)C

yste

ine/

ribos

e(2

,3,

6),

cyste

ine/

gluc

ose

(4),

cyste

ine/

rham

nose

(4),

glut

athi

one/

ribos

e(3

),th

iam

ine

(3,

5),

thia

min

e/cy

stein

e(5

),cy

stein

e/IM

P(6

,7

),cy

stein

e/rib

ose-

5-P

(6)

22-

Furfu

rylth

iol

Sulfu

ry,r

oasty

,cof

fee-

like

0.01

(1)

Cys

tein

e/rib

ose

(2,

3,

6),

cyste

ine/

gluc

ose

(4),

cyste

ine/

rham

nose

(4),

glut

athi

one/

ribos

e(3

),th

iam

ine

(3),

cyste

ine/

IMP

(6,

7),

cyste

ine/

ribos

e-5-

P(6

)

3M

erca

ptoa

ceta

ldeh

yde

Cab

bage

-like

—Th

iam

ine/

cyste

ine

(5)

4M

erca

pto-

2-pr

opan

one

Sulfu

ry,p

utrid

—C

yste

ine/

IMP

(7)

53-

Mer

capt

o-2-

buta

none

Sulfu

ry,c

atty

3.0

(1)

Cys

tein

e/rib

ose

(2,

6),

cyste

ine/

gluc

ose

(4),

cyste

ine/

rham

nose

(4),

thia

min

e/cy

stein

e(5

),cy

stein

e/IM

P(6

,7

),cy

stein

e/rib

ose-

5-P

(6)

63/

2-M

erca

pto-

2/3-

pent

anon

eSu

lfury

,cat

ty0.

7(1

)C

yste

ine/

ribos

e(2

,3,

6),

cyste

ine/

gluc

ose

(4),

cyste

ine/

rham

nose

(4),

glut

athi

one/

ribos

e(3

),th

iam

ine

(3,

5),

thia

min

e/cy

stein

e(5

),cy

stein

e/IM

P(6

),cy

stein

e/rib

ose-

5-P

(6)

7M

ethi

onal

Coo

ked

pota

to-li

ke0.

2(8

)M

ethi

onin

e/rib

ose

(9),

thia

min

e/m

ethi

onin

e(5

)

82-

Met

hyl-3

-thio

phen

ethi

olM

eaty

,sul

fury

0.02

(1)

Cys

tein

e/rib

ose

(2,

6),

cyste

ine/

IMP

(6),

cyste

ine/

ribos

e-5-

P(6

),th

iam

ine/

met

hion

ine

(5)

92-

Met

hyl-4

,5-d

ihyd

ro-3

-fura

nthi

olM

eaty

,sul

fury

—Th

iam

ine

(3,

5),

thia

min

e/cy

stein

e(5

)

10Bi

s(2-

met

hyl-3

-fury

l)di

sulfi

deM

eaty

,sul

fury

0.00

002

(10

)C

yste

ine/

ribos

e(2

,3,

6),

glut

athi

one/

ribos

e(3

),th

iam

ine

(5),

cyste

ine/

IMP

(6),

cyste

ine/

ribos

e-5-

P(6

)

112-

Then

ylth

iol

Sulfu

ry,r

oasty

0.04

2(1

)C

yste

ine/

ribos

e(2

,6

),cy

stein

e/gl

ucos

e(4

),th

iam

ine/

cyste

ine

(5),

cyste

ine/

IMP

(6,

7),

cyste

ine/

ribos

e-5-

P(6

)

(Con

tinue

d)



Table

3.3

(Con

tinue

d)

Odour

thre

shold

No.

Com

pound

Odour

des

crip

tion

(�g/k

g)

inw

ate

rbD

etec

ted

inb

125-

Met

hyl-2

-furfu

rylth

iol

Sulfu

ry,r

oasty

0.04

8(1

)C

yste

ine/

rham

nose

(4)

135-

Met

hyl-2

-then

ylth

iol

Sulfu

ry,r

oasty

0.04

9(1

)C

yste

ine/

rham

nose

(4)

142-

(1-M

erca

ptoe

thyl

)fura

neSu

lfury

,bur

nt0.

022

(1)

Cys

tein

e/gl

ucos

e(4

)

152-

(1-M

erca

ptoe

thyl

)thio

phen

eSu

lfury

,bur

nt0.

038

(1)

Cys

tein

e/gl

ucos

e(4

)

162-

Met

hylte

trahy

drot

hiop

hen-

3-on

eSu

lfury

,bur

nt—

Cys

tein

e/rib

ose

(2,

3),

thia

min

e(5

),th

iam

ine/

cyste

ine

(5),

cyste

ine/

IMP

(6,

7),

cyste

ine/

ribos

e-5-

P(6

)

171-

(Met

hylth

io)e

than

ethi

olTh

iam

ine-

like,

mea

ty—

Thia

min

e/cy

stein

e(5

),th

iam

ine/

met

hion

ine

(5)

182-

Met

hyl-1

,3-d

ithio

lane

Sulfu

ry—

Thia

min

e/cy

stein

e(5

)

19H

ydro

gen

sulfi

deSu

lfury

,egg

-like

10(1

1)

Cys

tein

e/rib

ose

(2),

cyste

ine/

gluc

ose

(4),

cyste

ine/

rham

nose

(4)

20M

etha

neth

iol

Sulfu

ry,p

utrid

0.2

(12

)C

yste

ine/

ribos

e(2

),cy

stein

e/gl

ucos

e(4

),cy

stein

e/rh

amno

se(4

)

21Et

hane

thio

lSu

lfury

,put

rid—

Cys

tein

e/rib

ose

(2),

cyste

ine/

gluc

ose

(4),

cyste

ine/

rham

nose

(4)

222-

Ace

tyl-2

-thia

zolin

eRo

asty

,pop

corn

-like

1.0

(13

)C

yste

ine/

ribos

e(2

),cy

stein

e/gl

ucos

e(4

),cy

stein

e/rh

amno

se(4

)

235-

Ace

tyl-2

,3-d

ihyd

ro-1

,4-th

iazi

neRo

asty

,pop

corn

-like

1.25

(1)

Cys

tein

e/rib

ose

(2),

cyste

ine/

gluc

ose

(4),

cyste

ine/

rham

nose

(4)

244-

Hyd

roxy

-2,5

-met

hyl-3

(2H

)-th

ioph

enon

eC

aram

el-li

ke,s

wee

t,m

eaty

24.0

(1)

Cys

tein

e/gl

ucos

e(4

)



B:C

ompo

unds

cont

aini

ngox

ygen

254-

Hyd

roxy

-2,5

-dim

ethy

l-3(2

H)-f

uran

one

(Fur

aneo

l)C

aram

el,s

traw

berr

y-lik

e10

(14

)C

yste

ine/

ribos

e(2

),cy

stein

e/gl

ucos

e(4

),cy

stein

e/rh

amno

se(4

),pr

olin

e/gl

ucos

e(1

5),

prol

ine/

rham

nose

(16

),ly

sine

/rha

mno

se(1

6)

264-

Hyd

roxy

-5-m

ethy

l-3(2

H)-f

uran

one

(nor

fura

neol

)C

aram

el-li

ke,b

urnt

chic

ory

8500

(1)

Cys

tein

e/rib

ose

(2),

prol

ine/

xylo

se(1

6)

272-

Ethy

l-4-h

ydro

xy-5

-met

hyl-3

(2H

)-fu

rano

ne(h

omof

uran

eol)

Car

amel

-like

,sw

eet

1.15

(14

)C

yste

ine/

rham

nose

(4)

283-

Hyd

roxy

-2-m

ethy

l-4(4

H)-p

yran

one

(mal

tol)

Car

amel

-like

3500

0(1

7)

Serin

e/m

alto

se(1

8),

prol

ine/

lact

ose

(18

)

293-

Hyd

roxy

-4,5

-dim

ethy

l-2(5

H)-f

uran

one

(sot

olon

)Se

ason

ing-

like

0.3

(19

)C

yste

ine/

ribos

e(2

),cy

stein

e/gl

ucos

e(4

),cy

stein

e/rh

amno

se(4

)

303-

Hyd

roxy

-6-m

ethy

l-2(2

H)-p

yran

one

Seas

onin

g-lik

e15

.0(1

)C

yste

ine/

rham

nose

(4)

C:C

ompo

unds

cont

aini

ngni

troge

n

312,

3-D

ieth

yl-5

-met

hylp

yraz

ine

Earth

y,ro

asty

0.09

(20

)C

yste

ine/

gluc

ose

(4),

cyste

ine/

rham

nose

(4)

322-

Ace

tyl-1

-pyr

rolin

eRo

asty

,pop

corn

-like

0.1

(21

)Pr

olin

e/gl

ucos

e(1

5)

336-

Ace

tylte

trahy

drop

yrid

ine

Roas

ty,b

urnt

,car

amel

-like

1.6

(22

)Pr

olin

e/gl

ucos

e(1

5)

342-

Ace

tylp

yrid

ine

Roas

ty,c

aram

el-li

ke19

(23

)Pr

olin

e/gl

ucos

e(1

5)

aTh

ese

nsor

ysi

gnifi

canc

eof

the

arom

aco

mpo

unds

was

asse

ssed

byqu

antit

ativ

eda

taor

byG

C-O

tech

niqu

es.

b1

,Hof

man

n(1

995)

;2,H

ofm

ann

and

Schi

eber

le(1

995)

;3,G

asse

r(19

90);

4,H

ofm

ann

and

Schi

eber

le(1

997)

;5,G

unte

rtet

al.(

1996

);6

,Mot

tram

and

Nob

rega

(199

8);7

,Zha

ngan

dH

o(1

991)

;8,G

uada

gnie

tal.

(197

2);9

,Mey

nier

and

Mot

tram

(199

5);1

0,B

utte

ryet

al.(

1984

);1

1,P

ippe

nan

dM

ecch

i(19

69);

12

,Gut

han

dG

rosc

h(1

994)

;13

,Cer

nyan

dG

rosc

h(1

993)

;14

,Sem

mel

roch

etal

.(19

95);

15

,Rob

erts

and

Acr

ee(1

994)

;16

,Dec

nop

etal

.(19

90);

17

,Pitt

etet

al.(

1970

);1

8,F

icke

rt(1

999)

;19

,Ker

lera

ndG

rosc

h(1

997)

;20

,G

rosc

h(2

001)

;21

,But

tery

etal

.(19

83);

22

,But

tery

and

Ling

(199

5);2

3,T

eran

ishi

etal

.(19

75).



Furaneol, which is also a key ingredient of caramel-like process flavours, can be ef-ficiently prepared by reacting rhamnose or other 6-deoxyhexoses with lysine, proline orhydroxyproline (Decnop et al., 1990). Another important caramel-like aroma compound,3-hydroxy-2-methyl-4(4H)-pyranone (maltol), is a key substance of process flavours that areprepared from disaccharides (maltose or lactose) and proline or serine (Fickert, 1999).

In thermally treated solutions of proline and glucose or fructose, ACP and ACTP havebeen evaluated as important aroma compounds, contributing a roasty, popcorn-like odour tothese process flavours (Roberts and Acree, 1994; Schieberle, 1995). ACP and ACTP have alsobeen reported to be character-impact compounds of various thermally treated cereal products(see also Section 3.3.1). In addition, Roberts and Acree (1994) showed that Furaneol and2-acetylpyridine contribute to the flavour of the proline/glucose model system.

3.4 PREPARATION OF PROCESS FLAVOURS

3.4.1 General aspects

In Europe, flavourings that are obtained by thermal treatment of a reducing sugar and a food-grade nitrogen source such as amino acids, peptides, food proteins, hydrolysed vegetableproteins (HVPs) and yeasts are referred to as thermal process flavourings (official termi-nology revised by EU in 2008; see Chapter 2 for more details). These products have beendesignated as a separate class of flavours and are identified as complex mixtures that havebeen converted to flavours by heat processing. In the US, the term ‘process flavours’ doesnot exist in regulatory terms. Maillard reaction flavours are considered natural or artificialflavours, depending on whether the starting materials and process are considered natural.The International Organisation of the Flavour Industry (IOFI) has established a guideline formanufacturers of process flavours. This guideline is part of the IOFI’s Code of Practice anddefines the types of raw materials and general reaction conditions (for instance, a maximumtemperature/time treatment of 180◦C/15 minutes, pH ≤8) (reviewed by Manley, 1995). TheUS Department of Agriculture, however, did not set a guideline for manufacture but estab-lished labelling criteria for materials used to produce a process flavour (Lin, 1995). Althoughprocessed flavours that are prepared according to the IOFI guidelines were considered GRAS(generally recognised as safe) in the US in 1995, the regulatory status of Maillard reactionflavours still lacks clarity with respect to the GRAS specification. This is due to lack ofinformation on whether processed flavours contain heterocyclic amines in amounts sufficientto affect their safe use in foods.

Maillard reaction technology is commonly used by the flavour industry to produce com-plex building blocks that provide similar aroma and taste properties to thermally treatedfoodstuffs such as meat, chocolate, coffee, caramel, popcorn or bread. Although flavour for-mation during the Maillard reaction is quantitatively a minor pathway, process flavours arevery important to the flavour industry. This is because these complex blocks exhibit uniqueflavour qualities, are difficult to copy and are relatively cheap, being based on low productioncosts and high flavour potency of the aroma compounds formed.

3.4.2 Factors influencing flavour formation

The factors that influence flavour formation and, thus, the sensory properties of processflavours, are the type of sugar and amino acid, pH, reaction media, water activity as well as



Table 3.4 Flavour types of processed sugar–amino acid model mixtures.

Amino Temperature FlavourSugar acid (◦C) description Referencesa

Glucose Cysteine 100–140 Meaty, beefy 1, 6Ribose Cysteine 100 Meaty, roast beef 1, 2, 6Ascorbic acid Threonine 140 Beef extract, meaty 1Ascorbic acid Cysteine 140 Chicken 1Glucose Serine or glutamine or

tyrosine100–220 Chocolate 1

Glucose Leucine 100 Chocolate 3Glucose Threonine 100 Chocolate 3Glucose Phenylalanine 100–140 Floral, chocolate 1Ribose orxylose

Threonine 140 Almond, marzipan 1

Glucose Proline 100–140 Nutty 1Glucose Proline or

hydroxyproline180 Bread, baked 3, 6

Glucose Alanine 100–220 Caramel 1, 6Glucose Lysine 110–120 Caramel 4Xylose Lysine 100 Caramel, buttery 5Ribose Lysine 140 Toast 1Glucose Valine 100 Rye bread 3Glucose Arginine 100 Popcorn 3Glucose Methionine 100–140 Cooked potatoes 1, 3Glucose Isoleucine 100 Celery 1Glucose Glutamine or

asparagine—b Nutty 6

a 1, Lane and Nursten (1983); 2, Morton et al. (1960); 3, Herz and Schallenberger (1960); 4, McKenna (1988); 5,Apriyantono and Ames (1990); 6, Yaylayan et al. (1994).b Microwave heating (640 W for 2–4 minutes).

temperature and time (see reviews by Shibamoto, 1983; Reineccius, 1990). In general, thesensory quality of a process flavour is less influenced by the type of sugar than by the aminoacid. Several authors (Herz and Schallenberger, 1960; Lane and Nursten, 1983; Yaylayanet al., 1994) studied the variety of odours produced in Maillard model systems, comprisingtwo or more components in reaction systems containing various sugars (or ascorbic acid) witheach of the protein-derived amino acids. The flavour characters of some of these processedsugar–amino acid model mixtures are summarised in Table 3.4.

Cysteine is the favoured amino acid to produce meat-like flavours, both on heating withreducing sugars or alone (Lane and Nursten, 1983). These authors also obtained chicken andbeef aromas by reacting, respectively, cysteine and threonine with ascorbic acid. Chocolateflavours can be prepared by heating glucose with amino acids such as serine, glutamine,tyrosine, leucine, threonine or phenylalanine (Herz and Schallenberger, 1960; Lane andNursten, 1983). Phenylalanine also gives rise to a floral aroma (in reaction with glucose oralone), whereas threonine yields nutty aromas when reacted with ribose or xylose. Prolineis the favoured amino acid for the production of bread-like and baked flavours (Lane andNursten, 1983; Yaylayan et al., 1994). However, Schieberle (1992a) showed that the yeast-derived amino acids ornithine and citrulline are even more effective precursors for ACP,which is a key aroma compound of bread crust. Herz and Schallenberger (1960) reported thegeneration of rye bread and popcorn aromas by heating valine or arginine with glucose. In



addition, alanine and lysine were found to give caramel flavours, and glutamine and arginine,nutty flavours.

Besides the type of sugar and amino acid, pH is another important factor determiningaroma of process flavours. It is well known to the flavour industry that meat flavours arepreferably prepared at low pH (4–5.5), whereas roast and caramel flavours are obtainedunder neutral or slightly basic conditions. Madruga and Mottram (1995) as well as Hofmannand Schieberle (1998a) showed that important sulfur-containing compounds in meat, suchas MFT, 2-furfurylthiol and 2-methyl-3-(methyldithio)furan, are preferably formed at a pHof 3–4. Sensorial evaluations of thermally treated model mixtures of ribose or 5′-IMP andcysteine revealed that the highest scores for boiled meat character were obtained when thereactions were carried out at pH 4.5 (ribose) and 3 (5′-IMP) (Madruga and Mottram, 1998).

Reaction media and water activity of the Maillard reaction systems are additional fac-tors that influence aroma generation. Besides buffered aqueous solutions, solvents such aspropylene glycol, glycerol, triacetin or fats and oils, as well as their emulsions or mixtureswith water, are used. Vauthey et al. (1998), for example, filed a patent on the generation ofroast chicken aroma using a cubic phase system. This system was prepared by introducinga melted monoglyceride (saturated in C16 and C18) into an aqueous phosphate buffer solu-tion. Compared to the same reaction in phosphate buffer, the flavour formed in the cubicsystem was more intense, corresponding to higher amounts of sulfur compounds such asMFT. Shu and Ho (1989) investigated the reaction of cysteine and 4-hydroxy-2,5-dimethyl-3(2H)-furanone in varying proportions of water and glycerol. They found that a superiorroasted/meaty character was obtained in the aqueous system. The influence of the water ac-tivity on pyrazine formation during Maillard reaction was studied by Leathy and Reineccius(1989a). The authors observed that pyrazine formation was optimal at an Aw of about 0.75.

Schieberle and Hofmann (1998) compared the character-impact compounds of cysteine-based process flavours formed in aqueous solution and under dry heating conditions. In acysteine/ribose model system, dry heating yielded higher amounts of key odourants withroasty notes such as 2-furfurylthiol, 2-acetyl-2-thiazoline and 2-propionyl-2-thiazoline aswell as 2-ethyl- and 2-ethenyl-3,5-dimethylpyrazine, whereas the meat-like sulfur com-pounds MFT and MP were found in comparable or lower concentrations, respectively. Theirstudy also revealed that the amounts of the 3-deoxyosone-derived compounds 2-furfuraland 5-methylfuran-2-aldehyde were significantly higher in the dry-heated model systems,whereas the formation of the 1-deoxyosone-derived 4-hydroxy-2,5-dimethyl-3(2H)-furanone(Furaneol) was enhanced in aqueous solution. An explanation for this finding could be that,under dry heating conditions, caramelisation processes are favoured relative to Maillardreaction. In caramelisation processes, 1,2-enolisation of the sugar molecule is preferredover 2,3-enolisation, leading to the formation of high amounts of 3-deoxyosones (Kroh,1994).

Apart from the precursor composition (e.g. availability of amino acids as reactants), pHand the reaction medium, other reaction parameters such as the presence of catalysts (e.g.phosphates) as well as temperature and time have a major effect on the sensory propertiesof process flavourings. Many of these parameters have extensively been studied for thegeneration of the caramel-like smelling Furaneol from 6-deoxyhexoses (Schieberle, 1992b;Havela-Toledo et al., 1997, 1999; Hofmann and Schieberle, 1997, 1998b; Schieberle andHofmann, 2002). For example, the yield of Furaneol from rhamnose (pH 7, 150◦C, 45minutes) was increased about 40 times when the malonate buffer was replaced with phos-phate buffer (Schieberle, 1992b). Even a higher increase of its yield (70-fold) was observedwhen the pH of rhamnose/cysteine system (145◦C, 20 minutes) was increased from 3 to 7



(Schieberle and Hofmann, 2002). Recently, Illmann et al. (2009) used a fractional factorialdesign to identify critical reaction parameters affecting kinetics of Furaneol generation fromrhamnose under cooking conditions (120◦C). The importance of the reaction parameterswas found to decrease in the following order: phosphate concentration > concentration ofprecursors � pH � rhamnose to lysine ratio. The experimental design approach achievedvery high yields of Furaneol (about 40 mol%). The type of amino acid was also shownto affect the yield of Furaneol from rhamnose. Lysine was most efficient in generating thecaramel-like odourant, followed by alanine, serine, glycine and threonine. On the other hand,much lower yields were obtained in the presence of proline and especially cysteine (Davideket al., 2009).

The generation of 3-hydroxy-2-methyl-4(4H)-pyranone (maltol), another importantcaramel-like aroma compound, was shown to strongly depend on the reaction media. Ratherlow yields of maltol were obtained when lactose was heated with proline in aqueous systems(13.5 mmol/mol lactose) or when the precursors were dry heated (7.2 mmol/mol lactose).However, the yield of maltol increased significantly when water was replaced by propyleneglycol (70 mmol/mol lactose; Cerny, 2003).

In addition, the knowledge of the reaction kinetics of aroma compounds helps to explainthe influence of temperature and time on their formation (Reineccius, 1990). Althoughflavour formation is a multi-step reaction sequence, Arrhenius kinetics have been found todescribe flavour formation well in model and real food systems. Stahl and Parliment (1994)used an ingenious device to obtain clear time/temperature conditions for model systems anddetermined the activation energies of flavour compounds. Leathy and Reineccius (1989b)showed that the sensory quality of a product is less influenced by temperature/time in a modelsystem that is designed to result in similar key aroma compounds (e.g. pyrazines). This can beexplained by the fact that these compounds have similar activation energies. However, theirstudy focused on dialkylpyrazines and did not consider the more potent trialkylpyrazines,which were suggested to have different formation pathways (Amrani-Hemaimi et al., 1995;Schieberle and Hofmann, 1998). As a result, flavour generation during Maillard reaction isin most cases strongly influenced by temperature and time, also because Maillard reactionflavours are complex mixtures of different classes of key aroma compounds. Lee (1995)proposed a different approach, on the basis of differential equations, to describe Maillardkinetics. The model produced was able to simulate the Maillard reaction not just as a functionof time and temperature but also as a function of reactant concentrations and pH. From thissimulation, the relative amounts of reactants could be plotted against time. Examples giveninclude the concentrations of aldose and Amadori compounds over a 10-hour reaction and thetime profiles of enolisation compounds from both the 1,2- and 2,3-pathways under definedpH conditions.

The use of stable intermediates as flavour precursors and/or the application of multi-step reactions are another way of optimising processing conditions on the basis of therequired application. Blank et al. (2003a,b) compared the generation of odourants in Maillardreaction systems containing glucose and proline (Glc/Pro) or the corresponding Amadoricompound fructosyl-proline (Fru-Pro). The major odourants found in both systems weresimilar and included ACP, 4-hydroxy-2,5-dimethyl-3(2H)furanone (Furaneol), acetic acid, 3-hydroxy-4,5-dimethyl-2(5H)furanone (sotolon), 2,3-butanedione and ACTP. However, theirconcentrations as well as their contributions to the final flavour differed. For example, theformation of Furaneol was favoured from Fru-Pro, namely at pH 6 and 7. On the other hand,the reaction system Glc/Pro gave rise to relative high yields of ACP and ACTP. Although bothroasted-smelling popcorn odourants showed high sensory relevance in both reaction systems,



they dominated especially the aroma of the Glc/Pro process flavouring. These results alsoindicate that there is no benefit in using Amadori compound for the formation ACTP and ACP.

The following sections give a survey of the patent literature in the area of savoury andsweet process flavours. Emphasis is given to meat-like process flavours.

3.4.3 Savoury process flavours

The great majority of patents based on Maillard reaction technology have been directed to theproduction of meat-like process flavours. Most of these reaction flavours indicate cysteineand thiamine as the essential sulfur-containing precursor compounds. In 1960, the basicconcept of Maillard flavour technology was beginning to emerge with the Unilever patent ofMorton et al. (1960). The authors disclosed Maillard processes for the production of cookedbeef and pork flavours by reacting ribose or mixtures of ribose and glucose with cysteineand either additional amino acids or deflavoured protein hydrolysates from cod fish flesh,casein, groundnut or soya. The group of additional amino acids was consisted of �-alanine,glutamic acid, glycine, �-alanine, threonine, histidine, lysine, leucine, serine and valine. Allflavours were prepared in water at a pH between 3 and 6 and a temperature around 130◦C.Using similar reaction conditions, May and Morton (1960) and May (1961) also preparedmeat flavours through reacting cysteine with glyceraldehyde or furfural in combination withdeflavoured cod fish hydrolysates or amino acid mixtures.

Jaeggi (1973) patented processes for boiled and roasted beef flavours, which were alsobased on the reaction of ribose and cysteine. However, the inventor used methionine and pro-line as additional amino acids and carried out the reaction in glycerol or groundnut oil insteadof water. Methionine as the sole sulfur source was also reported to result in beef flavouringswhen reacted with xylose and cysteine-free hydrolysed plant protein (Van Pottelsberghede la Potterie, 1973). Tandy (1985) prepared process flavourings with white chicken meatcharacter using leucine and cysteine in combination with the reducing sugars arabinose andglucose. In addition, he found that the use of rhamnose instead of glucose (as well as theaddition of serine) provided a more aromatic, characteristic white meat chicken flavour.

International Flavors and Fragrances (IFF) filed a number of patents on the preparationof chicken, beef and pork flavours using cysteine in combination with thiamine, often incarbohydrate-free systems. This demonstrates that thiamine is capable of replacing carbo-hydrates by providing similar intermediates and aroma compounds as found in Maillardsystems. Intense beef flavours, for example, were obtained by refluxing cysteine, thiamineand carbohydrate-free vegetable protein hydrolysate (HVP) in water or water–ethanol mix-tures (IFF, 1965). The addition of beef tallow was found to result in a beef flavour with a‘pan-dripping’ character. In addition, chicken flavourings were prepared by heating cysteine,thiamine and HVP in combination with other ingredients such as �-alanine, glycine andascorbic acid, whereas pork flavours required the addition of methionine and lard. Similarprocesses for chicken, beef and pork flavours were disclosed in other IFF patents (IFF, 1967;Giacino, 1968a, 1969; Katz and Evers, 1973). Chicken aroma was found to be improved, forexample, by adding diacetyl and hexanal (Giacino, 1968a) or mercaptoalkanones (Katz andEvers, 1973) to the processed flavours. Dihydroxyacetone, pyruvic acid or pyruvic aldehydein combination with thiamine and HVPs were claimed to result in beef flavourings withimproved cooked note (Giacino, 1968b).

Kerscher (2000) investigated the analytical assessment of the species-specific characterof beef, chicken and pork. His results (see Section 3.3.1) cannot explain the choice of the



ingredients for the preparation of chicken, beef and pork flavours that are disclosed in the IFFpatents mentioned above, but are in agreement with the findings of Chen and Tandy (1988).The authors developed species-specific beef and chicken flavourings through oxidation ofoleic and linoleic acids, respectively. Their processes involved heat treatment of oleic orlinoleic acid in the presence of air at high temperatures of about 300◦C as well as trappingof the resulting aroma fraction in cold traps. The authors also claimed flavour blocks thatresembled roast, grilled, bloody and braised beef in different fractions of the distillate ofoxidised oleic acid.

In terms of alternative sulfur sources to cysteine and thiamine, Giacino (1970) found thatprocess flavours with similar characters were obtained when cysteine was replaced by taurine.Patents of the Corn Products Company also describe the production of Maillard reactionflavours using taurine in combination with HVPs and xylose (Corn Products Company,1969; Hack and Konigsdorf, 1969). From a scientific point of view, however, the finding thatcysteine can be replaced by taurine has to be questioned, because there are no studies thatreport the generation of important meat aroma compounds from taurine. In addition, Tai andHo (1997) could detect only trace amounts of volatile sulfur compounds in a Maillard modelsystem containing cysteinesulfinic acid and glucose. Broderick and Linteris (1960) usedderivatives of mercaptoacetaldehyde such as 2,5-dihydroxy-1,4-dithiane, diethyl- or dithio-acetals and hemimercaptals as precursors to impart meat-like flavour to canned simulatedmeat and vegetable products on sterilisation. The patent of Heyland (1977) involved theuse of hydrolysed onion, garlic and cabbage in combination with HVP, ribose and beef fatfor the preparation of beef flavours. Other flavour companies developed meat flavouringswith sulfur sources such as hydrogen, sodium or ammonium sulfides (Godman and Osborne,1972; Gunther, 1972), methionine (Van Pottelsberghe de la Potterie, 1972) or egg white(Theron et al., 1975).

Yeast extracts or yeast hydrolysates have traditionally been used either as precursorsfor the thermal generation of meat flavourings or as taste-enhancing ingredients, in blendswith process flavours. The advantage of yeast extracts is that they are a relatively cheap,natural source of amino acids and thiamine. In addition, their high content of glutamate and5′-ribonucleotides, particularly inosine 5′-monophosphate and guanosine 5′-monophosphate,provides complexity, body and flavour enhancement. Nestle, for example, filed several patentsin which the preparation of beef and chicken process flavours using yeast extracts wasdisclosed (Nestle, 1966; Rolli et al., 1988; Cerny, 1995). Cerny (1995) also developedbouillon flavours using yeast cream, which is enriched in hydrogen sulfide. Such a yeastcream was obtained by incubating baker’s yeast with elemental sulfur.

In order to obtain complete meat flavouring products, process flavours are blended withseveral other ingredients, which provide aroma (e.g. compounded aroma blocks referred toas top notes or topnote flavours), taste, taste enhancement, mouth feel and body (referredto as base notes). Besides topnote flavours, yeast extracts, hydrolysed vegetable proteinsand the monosodium salts of glutamate, inosinate and guanylate, ingredients such as onion,garlic, celery and/or caramel powder, animal or vegetable fats, gelatine and spices are oftenused in meat flavour compositions. The use of some yeast extracts, however, is limitedowing to their undesirable ‘yeasty’ character. Therefore, research groups have developedprocesses for manufacturing yeast hydrolysates with improved meat-like taste, in whichthe yeasty notes are absent. De Rooij and Hakkaart (1992), for example, improved themeaty character of yeast hydrolysates prepared from several yeast species by combiningthe enzymatic degradation of yeast cells with an additional fermentation step, which wascarried out using lactic acid-producing micro-organisms or additional yeasts. Hyoky et al.



(1996) developed a method for the production of yeast extracts in which undesirable bitterand yeasty flavour notes were removed. The authors evaluated several non-ionic and slightlybasic macroporous polymeric adsorbents as well as activated carbon for their ability to bindbitter and other undesirable flavouring substances of yeast hydrolysates, without bindingyeast peptides, amino acids or nucleotides. The best results were obtained with AmberliteXAD-16 and Amberlite XAD-765, which are a non-ionic styrene/divinylbenzene copolymerand a weakly basic phenolformaldehyde polymer, respectively.

3.4.4 Sweet process flavours

A few patents and articles on the generation of chocolate and caramel flavours are discussedhere. Rusoff (1958) prepared artificial chocolate flavours by heating partially hydrolysedproteins with sugars. The Maillard reactions were performed between protein hydrolysatesderived from casein, soy, wheat gluten or gelatine and mixtures of pentoses and hexoses.The reaction medium contained 30% water and the reaction temperature ranged between 130and 150◦C. These ‘base chocolate flavours’ were rounded off by adding ingredients such ascaffeine, theobromine and tannins before or after the heating process.

The need for hydrolysed proteins to generate chocolate flavours through Maillard reactionwas stressed by Rodel et al. (1988). The authors based their investigations on precursorstudies of Mohr et al. (1971, 1976), who found that only mixtures of peptide and free aminoacid fractions isolated from fermented raw cocoa beans developed cocoa aroma on thermaltreatment. The study of Rodel et al. (1988) covered a complete range of parameters suchas source of protein, rate of hydrolysis, source of enzyme, amount of sugar, water contentas well as temperature and time. The Maillard reaction flavours obtained were evaluatedorganoleptically and analytically. The authors revealed that gelatine that is enzymaticallyhydrolysed by more than 20% is an appropriate protein source for producing cocoa flavoursthrough Maillard reaction. The quality of the cocoa aroma was further affected by water andsugar contents, whereas the source of enzyme had no influence. A water content of at least5% and a sugar content of 20 g per 100 g hydrolysed protein gave a positive effect on flavour.In addition, temperature and time, which were the most sensitive parameters, were optimisedat 144◦C and 21 minutes.

Pittet and Seitz (1974) disclosed processes for the preparation of various flavours resem-bling chocolate, sweet corn, popcorn, bread, cracker and caramel toffee. The flavours wereprepared by heating cyclic enolones such as Furaneol, maltol or cyclotene with amino acidsin propylene glycol or glycerol. The temperatures ranged between 120 and 205◦C. Choco-late flavours, for example, were obtained when valine or leucine was reacted with maltolor Furaneol, whereas proline (in combination with Furaneol or maltol) yielded cracker-like,popcorn, sweet corn and bread aromas. In addition, caramel toffee and burnt sugar flavourswere established from proline and cyclotene or ethyl cyclotene. Gilmore (1988) also devel-oped a caramel butterscotch flavour by heating a mixture of sugar syrup and butter in thepresence of ammonia at a pH of 7 and a temperature of about 100◦C.

3.5 OUTLOOK

Flavour formation is a minor but important pathway within the complex cascade of chemicalreactions occurring during Maillard processes. A strong focus on the key flavour compounds(aroma and taste-active molecules), their precursors and reaction routes is required for the



optimisation of process flavours. In addition, the formation of undesirable molecules needsto be taken into account when it comes to the optimisation of processing parameters. Thebetter understanding of the influence of various process parameters on the formation of boththe key aroma compounds and their precursors is still a challenge for future research.

Experimental design and kinetic parameter estimations are good tools for limiting theamount of experiment required. In addition, the evaluation of important taste compoundsderived from the Maillard reaction, the elucidation of their formation pathways as well asthe understanding of how the interaction of aroma and taste compounds affects the sensorialquality of the flavours are certainly research areas that are worth investigating.

REFERENCES

Amrani-Hemaimi, M., Cerny, C. and Fay, L.B. (1995) Mechanisms of formation of alkylpyrazines in theMaillard reaction. J. Agric. Food Chem. 43, 2818–2822.

Anet, E.F.L.J. (1960) 3-Deoxyhexosones. J. Am. Chem. Soc. 82, 1502–1504.Anonymous (1979) Foodstuff with flavor precursor. Res. Disclosure 177–188, 17918.Apriyantono, A. and Ames, J.M. (1990) Volatile compounds produced on heating lysine with xylose. In:

Flavour Science and Technology (eds Y. Bessiere and A.F. Thomas), Wiley, Chichester, UK, pp. 117–120.

Beksan, E., Schieberle, P., Robert, F., Blank, I., Fay, L.B., Schlichtherle-Cerny, H. and Hofmann, T. (2003)Synthesis and sensory characterization of novel Umami-tasting glutamate glycoconjugates. J. Agric. FoodChem. 51, 5428–5436.

Belitz, H.-D., Grosch, W. and Schieberle, P. (2008) Food Chemistry, 4th edn, Springer, Berlin.Bel Rhild, B., Fleury, Y., Devaud, S., Fay, L.B., Blank, I. and Juillerat, M.A. (2002) Biogeneration of

roasted notes based on 2-acetyl-2-thiazoline and its precursor 2-(1-hydroxyethyl)-4,5-dihydrothiazole. In:Heteroatomic Aroma Compounds, ACS Symposium Series 826 (eds G.A. Reineccius and T.A. Reineccius),American Chemical Society, Washington, DC, pp. 179–190.

Blank, I., Devaud, S., Matthey-Doret, M., Pollien, P.., Robert, F. and Yeretzian, C. (2003b) Formation ofodour-active compounds in Maillard model systems based on proline. In: Flavour Research at the Dawnof the Twenty-First Century, Proceedings of the 10th Weurman Flavour Research Symposium (eds J.-L.Le Quere and P.X. Etievant), Lavoisier, Intercept, London, pp. 458–463.

Blank, I., Devaud, S., Matthey-Doret, M. and Robert, F. (2003a) Formation of odorants in Maillard modelsystems based on L-proline as affected by pH. J. Agric. Food Chem. 51, 3643–3650.

Blank, I., Sen, A. and Grosch, W. (1992) Aroma impact compounds of Arabica and Robusta coffee. Colloq.Sci. Int. Cafe 14, 117–129.

Broderick, J.J. and Linteris, L.L. (1960) Flavoring agents and process for imparting a meat-like flavor to anedible composition. US Patent 2955 041.

Buttery, R.G., Haddon, W.F., Seifert, R.M. and Turnbaugh, J.G. (1984) Thiamin odor and bis(2-methyl-3-furyl) disulfide. J. Agric. Food Chem. 32, 674–676.

Buttery, R.G. and Ling, L.C. (1995) Volatile flavor components of corn tortillas and related products.J. Agric. Food Chem. 43, 1878–1882.

Buttery, R.G., Ling, L.C. and Juliano, B.O. (1982) 2-Acetyl-1-pyrroline: an important aroma compound ofcooked rice. Chem. Ind. (London) 23, 958–959.

Buttery, R.G., Ling, L.C., Juliano, B.O. and Turnbaugh, J.G. (1983) Cooked rice aroma and 2-acetyl-1-pyrroline. J. Agric. Food Chem. 31, 823–826.

Cerny, C. (1995) Procede de transformation du soufre elementaire en sulfure. European Patent 778350.Cerny, C. (2003) Maltol generation from lactose in dry systems. In: Flavour Research at the Dawn of the

Twenty-First Century, Proceedings of the 10th Weurman Flavour Research Symposium (eds J.L. Le Quereand P.X. Etievant), Lavoisier, Intercept, London, pp. 544–547.

Cerny, C. (2007) Process flavourings. In: Flavourings. Production, Composition, Applications, Regulations(ed. H. Ziegler), Wiley-VCH, Weinheim, Germany, pp. 274–297.

Cerny, C. and Davidek, T. (2003) Formation of aroma compounds from ribose and cysteine during Maillardreaction. J. Agric. Food Chem. 51, 2714–2721.



Cerny, C. and Grosch, W. (1993) Quantification of character-impact odor compounds of roasted beef. Z.Lebensm.-Unter. Forsch. 196, 417–422.

Chan, F. and Reineccius, G.A. (1994) The reaction kinetics for the formation of isovaleraldehyde, 2-acetyl-1-pyrroline, di(H)di(OH)-6-methylpyranone, phenylacet-aldehyde, 5-methyl-2-phenyl-2-hexenal, and 2-acetylfuran in model systems. In: Maillard Reactions in Chemistry, Food and Health (eds T.P. Labuza,G.A. Reineccius, V. Monnier, J. O’Brien and J. Baynes), The Royal Society of Chemistry, Cambridge,pp. 131–139.

Chen, T.-K. and Tandy, J.S. (1988) Production of flavour. European Patent 295509.Corn Products Company (1969) Process for the production of seasonings with a flavor similar to meat extract.

GB Patent 1148449.Cremer, D.R. and Eichner, K. (2000) The influence of the pH value on the formation of Strecker aldehydes

in low moisture model systems and in plant powders. Eur. Food Res. Technol. 211, 247–251.Davidek, T., Clety, N., Aubin, S. and Blank, I. (2002) Degradation of the Amadori compound N-(1-deoxy-

D-fructos-1-yl)glycine in aqueous model systems. J. Agric. Food Chem. 50, 5472–5479.Davidek, T., Devaud, S., Robert, F. and Blank, I. (2006a) Sugar fragmentation in the Maillard reaction

cascade: formation of acetic acid by a hydrolytic �-dicarbonyl cleavage mechanism. J. Agric. FoodChem. 54, 6667–6676.

Davidek, T., Illmann, S., Gouezec, E., Rytz, A., Schuchmann, H.P. and Blank, I. (2009) Effect of reactionconditions on generation of 4-hydroxy-2,5-dimethyl-3(2H)-furanone from rhamnose. In: 12th WeurmanFlavor Research Symposium, Interlaken, Switzerland, July 1–4, 2008.

Davidek, T., Robert, F., Devaud, S., Arce Vera, F. and Blank, I. (2006b) Sugar fragmentation in the Maillardreaction cascade: formation of short-chain carboxylic acids by a new oxidative �-dicarbonyl cleavagepathway. J. Agric. Food Chem. 54 6677–6684.

Decnop, C., van Dort, J.M. and de Hey, J.T. (1990) Hydroxyfuranone preparation. European Patent 398417.De Rooij, J.F.M. and Hakkaart, M.J.J. (1992) A process for the preparation of food flavours. European Patent

191513.de Roos, K.B. (1992) Meat flavor generation from cysteine and sugars. In: Flavor Precursors: Thermal and

Enzymatic Conversions (eds R. Teranishi, G.R. Takeoka and M. Guntert), American Chemical Society,Washington, DC, pp. 203–216.

Doornbos, T., Haring, P.G.M. and Van Der Heijden, A. (1991) Use of diketone in food products. EuropeanPatent 293957.

Doornbos, T., Van Den Ouweland, G.A.M. and Tjan, S.B. (1981) Amadori compounds, derived from 6-deoxysugars, as flavour precursor. In: Progress in Food and Nutrition Science, Maillard Reactions in Food (ed.C. Eriksson), Pergamon Press, Oxford, pp. 57–63.

Eichner, K., Reutter, M. and Wittmann, R. (1994) Detection of Amadori compounds in heated foods. In:Thermally Generated Flavors: Maillard, Microwave, and Extrusion Processes (eds T.H. Parliament, M.J.Morello and R.J. McGorrin), American Chemical Society, Washington, DC, pp. 42–54.

Fickert, B. (1999) Untersuchungen zur Bildung von Aromastoffen bei der Malzung von Getreide. PhD thesis,Technical University of Munich, Garching, Germany.

Frank, O., Jezussek, M., Hofmann, T. (2003) Sensory activity, chemical structure, and synthesis of Mail-lard generated bitter-tasting 1-oxo-2,3-dihydro-1H-indolizinium-6-olates. J. Agric. Food Chem. 51,2693–2699.

Gasser, U. (1990) Fluchtige Verbindungen aus gekochtem Rind- und Huhnerfleisch, Fleischaromen undFleischextrakt. PhD thesis, Technical University of Munich, Garching, Germany.

Gasser, U. and Grosch, W. (1988) Identification of volatile flavor compounds with high aroma values fromcooked beef. Z. Lebensm.-Unters. Forsch. 186, 489–494.

Gasser, U. and Grosch, W. (1990) Primary odorants of chicken broth. A comparative study with meat brothsfrom ox and cow. Z. Lebensm.-Unters. Forsch. 190, 3–8.

Gasser, U. and Grosch, W. (1991) Aroma von gekochtem Schweinefleisch. Lebensmittelchemie 45, 15–16.Gi, U.-S. and Baltes, W. (1995) Model reactions on roast aroma formation, 15. Investigations on the formation

of pyrido[3,4-d]imidazoles during the Maillard reaction. J. Agric. Food Chem. 43, 2226–2230.Giacino, C. (1968a) Poultry flavor composition and process. US Patent 3394017.Giacino, C. (1968b) Product and process of reacting a proteinaceous substance with a sulfur-containing

compound to provide a meat-like flavor. US Patent 3394015.Giacino, C. (1969) Meat flavor compositions. GB Patent 1146337.Giacino, C. (1970) Meat flavor compositions. US Patent 3519437.Gilmore, L.T. (1988) Process for preparing a caramel butterscotch flavor syrup. US Patent 4753814.Godman, J.L. and Osborne, D.R.D. (1972) Edible products. GB Patent 1285568.



Grosch, W. (1994) Determination of potent odourants in foods by aroma extract dilution analysis (AEDA)and calculation of odour activity values (OAVs). Flavour Frag. J. 9, 147–158.

Grosch, W. (2001) Flavour of chemistry III: Coffee and volatile compounds. In: Coffee, Recent Developments(eds O. Vitzthum and R.J. Clarke), Blackwell Science, Oxford, pp. 68–89.

Grosch, W. and Schieberle, P. (1997) Flavor of cereal products – a review. Cereal Chem. 74(2), 91–97.Guadagni, D.G., Buttery, R.G. and Turnbaugh, J.G. (1972) Odor thresholds and similarity ratings of some

potato chip components. J. Sci. Food Agric. 23, 1435–1444.Guntert, M., Bertram, H.-J., Hopp, R., Silberzahn, W., Sommer, H. and Werkhoff, P. (1996) Thermal

generation of flavor compounds from thiamin and various amino acids. In: Recent Developments inFlavor and Fragrance Chemistry (eds R. Hopp and K. Mori), VCH-Verlagsgesellschaft, Weinheim,Germany, pp. 215–240.

Guntert, M., Bruing, J., Emberger, R., Hopp, R., Kopsel, M., Surburg, H. and Werkhoff, P. (1992) Thermallydegraded thiamin: a potent source of interesting flavor compounds. In: Flavor Precursors: Thermal andEnzymatic Conversions (eds R. Teranishi, G.R. Takeoka and M. Guntert), American Chemical Society,Washington, DC, pp. 140–163.

Gunther, R. (1972) Flavoring compositions produced by reacting hydrogen sulfide with a pentose. US Patent3642497.

Guth, H. and Grosch, W. (1993a) 12-Methyltridecanal, a species-specific odorant of stewed beef. Lebensm.Wissensch. Technol. 26, 171–177.

Guth, H. and Grosch, W. (1993b) Identification of potent odorants in static headspace samples of green andblack tea powders on the basis of aroma extract dilution analysis (AEDA). Flavour Frag. J. 8, 173–178.

Guth, H. and Grosch, W. (1994) Identification of the character impact odorants of stewed beef juice byinstrumental analysis and sensory studies. J. Agric. Food Chem. 42, 2862–2866.

Hack, A.W. and Konigsdorf, W. (1969) Production of seasonings with a flavor similar to meat extract. USPatent 3480447.

Havela-Toledo, E., Naim, M., Zehavi, U. and Rouseff, R.L. (1997) 4-Hydroxy-2,5-dimethyl-3(2H)-furanoneformation in buffers and model solutions of citrus juice. J. Agric. Food Chem. 45, 1314–1319.

Havela-Toledo, E., Naim, M., Zehavi, U. and Rouseff, R.L. (1999) Effect of L-cysteine and n-acetyl-L-cysteine on 4-hydroxy-2,5-dimethyl-3(2H)-furanone (Furaneol), 5-hydroxymethylfurfural and 5-methylfurfural formation and browning in buffer solutions containing rhamnose or glucose and arginine.J. Agric. Food Chem. 47, 4140–4145.

Henderson, S.K. and Nawar, W.W. (1981) Thermal interaction of linoleic acid and its esters with valine.J. Am. Oil Chem. Soc. 58, 632–635.

Herz, W.J. and Schallenberger, R.S. (1960) Some aromas produced by simple amino acid-sugar reaction.Food Res. 25, 491–494.

Heyland, S. (1977) Verfahren zur Herstellung eines Aromatisierungsmittels. German Patent 2546035.Hidalgo, F.J., Zamora, R. (2004) Strecker-type degradation produced by the lipid oxidation products 4,5-

epoxy-2-alkenals. J. Agric. Food Chem. 52, 7126–7131.Hidalgo, F.J. and Zamora, R. (2005) Interplay between the Maillard reaction and lipid peroxidation in

biochemical systems. Ann. N. Y. Acad. Sci. 1043, 319–326.Hidalgo, F. J. and Zamora, R. (2007) Conversion of phenylalanine into styrene by 2,4-decadienal in model

systems. J. Agric. Food Chem. 55, 4902–4906.Hodge, J.E. (1953) Chemistry of browning reactions in model systems. J. Agric. Food Chem. 1, 928–943.Hofmann, T. (1995) Charakterisierung intensiver Geruchsstoffe in Kohlenhydrat/Cystein-Modellreaktionen

und Klarung von Bildungswegen. PhD thesis, Technical University of Munich, Garching, Germany.Hofmann, T. (1999) Quantitative studies on the role of browning precursors in the Maillard reaction of

pentoses and hexoses with L-alanine. Eur. Food Res. Technol. 209, 113–121.Hofmann, T. (2005) Taste-active Maillard reaction products: the ‘tasty’ world of nonvolatile Maillard reaction

products. Ann. N. Y. Acad. Sci. 1043, 20–29.Hofmann, T., Munch, P. and Schieberle, P. (2000) Quantitative model studies on the formation of aroma-active

aldehydes and acids by Strecker-type reactions. J. Agric. Food Chem. 48, 434–440.Hofmann, T., Ottinger, H., Frank, O., Soldo, T., Cerny, C., Robert, F. and Blank I. (2001) Use of alpha-keto

enamine derivatives as cooling ingredients. Eur. Pat. Appl., EP 1157617 A2.Hofmann, T. and Schieberle, P. (1995) Evaluation of the key odorants in a thermally-treated solution of ribose

and cysteine by aroma extract dilution techniques. J. Agric. Food Chem. 43, 2187–2194.Hofmann, T. and Schieberle, P. (1997) Identification of potent aroma compounds in thermally-treated mixtures

of glucose/cysteine and rhamnose/cysteine using aroma extract dilution techniques. J. Agric. Food Chem.45, 898–906.



Hofmann, T. and Schieberle, P. (1998a) Quantitative model studies on the effectiveness of different precursorsystems in the formation of the intense food odorants 2-furfurylthiol and 2-methyl-3-furanthiol. J. Agric.Food Chem. 46, 235–241.

Hofmann, T. and Schieberle, P. (1998b) Identification of key aroma compounds generated from cysteine andcarbohydrates under roasting conditions. Z. Lebensm.-Unters. Forsch. 207, 229–236.

Hofmann, T. and Schieberle, P. (2000a) Formation of aroma-active Strecker-aldehydes by a direct oxidativedegradation of Amadori compounds. J. Agric. Food Chem. 48, 4301–4305.

Hofmann, T. and Schieberle, P. (2000b) Acetylformoin – an important progenitor of 4-hydroxy-2,5-dimethyl-3(2H)-furanone and 2-acetyltetrahydropyridine during thermal food processing. In: Flavour 2000 –Perception, Release, Evaluation, Formation, Acceptance, Nutrition/Health (ed. M. Rothe), Eigenverlag,Bergholz-Rehbrucke, Germany, pp. 311–322.

Hyoky, G., Sarkki, M.-L. and Tylli, M. (1996) Method of producing a yeast extract usable in foodstuffs,where undesirable flavouring agents in the extract have been removed. World Patent 96/38057.

IFF, International Flavors & Fragrances, Inc. (1965) Composition and process for preparation thereof. GBPatent 1099711.

IFF, International Flavors & Fragrances, Inc. (1967) Flavoring compositions. GB Patent 1069104.Illmann, S., Davidek, T., Gouezec, E., Rytz, A., Schuchmann, H.P. and Blank, I. (2009) Generation of 4-

hydroxy-2,5-dimethyl-3(2H)-furanone from rhamnose as affected by reaction parameters – experimentaldesign approach. J. Agric. Food Chem. 57, 2889–2895.

Ishizu, A., Lindberg, B. and Theander, O. (1967) 1-Deoxy-D-erythro-2,3-hexodiulose, an intermediate in theformation of D-glucosaccharinic acid. Carbohydr. Res. 5, 329–334.

Jaeggi, K. (1973) Process for the manufacture of meat flavors. US Patent 3761287.Katz, I. and Evers, W.J. (1973) Flavoring compositions and processes utilizing alpha-ketothiols. US Patent

3773524.Kerler, J. (1996) Objektivierung eines Aromafehlers (Warmed-over Flavour) bei aufgewarmtem Huhner-und

Rindfleisch. PhD thesis, Technical University of Munich, Garching, Germany.Kerler, J. and Grosch, W. (1996) Odorants contributing to warmed-over flavor (WOF) of refrigerated cooked

beef. J. Food Sci. 61(6), 1271–1274, 1284.Kerler, J. and Grosch, W. (1997) Character impact odorants of boiled chicken: changes during refrigerated

storage and reheating. Z. Lebensm.-Unters. Forsch. 205, 232–238.Kerscher, R. (2000) Objektivierung tierartspezifischer Aromaunterschiede bei erhitztem Fleisch. PhD thesis,

Technical University of Munich, Garching, Germany.Kerscher, R. and Grosch, W. (1997) Comparative evaluation of potent odorants of boiled beef by aroma

extract dilution and concentration analysis. Z. Lebensm.-Unters. Forsch. 204, 3–6.Kerscher, R. and Grosch, W. (1998) Quantification of 2-methyl-3-furanthiol, 2-furfurylthiol, 3-mercapto-2-

pentanone, and 2-mercapto-3-pentanone in heated meat. J. Agric. Food Chem. 46, 1954–1958.Keyhani, A. and Yaylayan, V.A. (1996) Elucidation of the mechanism of pyrazidone formation in glycine

model systems using labeled sugars and amino acids. J. Agric. Food Chem. 44, 2511–2516.Kim, Y.-S., Hartman, T.G., Ho, C.-T. (1996) Formation of 2-pentylpyridine from the thermal interaction of

amino acids and 2,4-decadienal. J. Agric. Food Chem. 44, 3906–3908.Kroh, L.W. (1994) Caramelisation in food and beverages. Food Chem. 51, 373–379.Lane, M.J. and Nursten, H.E. (1983) The variety of odors produced in Maillard model systems and how they

are influenced by reaction conditions. In: Maillard Reaction in Food and Nutrition (eds G.R. Waller andM.S. Feather), American Chemical Society, Washington, DC, pp. 141–158.

Leathy, M.M. and Reineccius, G.A. (1989a) Kinetics on the formation of alkylpyrazines: effect of pH andwater activity. In: Thermal Generation of Aromas (eds T.H. Parliament, R.J. McGorrin and C.-T. Ho),American Chemical Society, Washington, DC, pp. 196–208.

Leathy, M.M. and Reineccius, G.A. (1989b) Kinetics on the formation of alkylpyrazines: effect of amino acidtype and type of sugar. In: Flavor Chemistry (eds R. Teranishi, R.G. Buttery and F. Shahidi), AmericanChemical Society, Washington, DC, pp. 76–91.

Ledl, F. and Schleicher, E. (1990) New aspects of the Maillard reaction in foods and in the human body.Angew. Chem. Int. Ed. Engl. 29, 565–706.

Lee, P.S. (1995) Modeling the Maillard reaction – a computer-simulation and a discussion of its applicationto Maillard reaction analysis and design. In: Flavor Technology: Physical Chemistry, Modification,and Process (eds C.-T. Ho, C.T. Tan and C.H. Tong), American Chemical Society, Washington, DC,pp. 74–97.

Limacher, A., Kerler, J., Davidek, T., Schalzried, F. and Blank, I. (2008) Formation of furan and methylfuranby Maillard-type reactions in model systems and food. J. Agric. Food Chem. 56, 3639–3647.



Lin, L.J. (1995) Regulatory status of Maillard reaction flavors. In: Savory Flavors (ed. T.W. Nagodawithana),Esteekay Associates, Wisconsin, pp. 7–15.

MacLeod, G. and Ames, J. (1986) 2-Methyl-3-(methylthio)furan: a meaty character impact compound fromcooked beef. Chem. Ind. 3, 175–177.

Madruga, M.S. and Mottram, D.S. (1995) The effect of pH on the formation of Maillard-derived aromavolatiles using a cooked meat system. J. Sci. Food Agric. 68, 305–310.

Madruga, M.S. and Mottram, D.S. (1998) Sensory analysis of a model system using 5′-IMP and cysteine atdifferent pH. Cienc. Tecnol. Aliment. 18(4), 397–404.

Manley, C.H. (1995) Process flavors and precursor systems. In: Savory Flavors (ed. T.W. Nagodawithana),Esteekay Associates, Wisconsin, pp. 16–25.

May, C.G. (1961) Flavouring substances and their preparation. GB Patent 858333.May, C.G. and Morton, I.D. (1960) Process for the preparation of a meat flavor. US Patent 2934436.McKenna, A.B. (1988) Influence of heating on the formation of caramel flavour, brown colour and hydrox-

ymethylfurfural in reconstituted concentrated skim milk. New Zeal. J. Dairy Sci. Technol. 23, 363–372.Meynier, A. and Mottram, D.S. (1995) The effect of pH on the formation of volatile compounds in meat-

related model systems. Food Chem. 52, 361–366.Milo, C. and Grosch, W. (1993) Changes in the odorants of boiled trout (Salmo fario) as affected by the

storage of the raw material. J. Agric. Food Chem. 41, 2076–2081.Milo, C. and Grosch, W. (1995) Detection of odor defects in boiled cod and trout by gas chromato-

graphy–olfactometry of headspace samples. J. Agric. Food Chem. 43, 2076–2081.Milo, C. and Grosch, W. (1996) Changes in the odorants of boiled salmon and cod as affected by the storage

of the raw material. J. Agric. Food Chem. 44, 459–462.Mohr, W., Landschreiber, E. and Severin, T. (1976) Zur Spezifitat des Kakaoaromas. Fett. Wiss. Technol.

78(2), 88–95.Mohr, W., Rohrle, M. and Severin, T. (1971) Uber die Bildung des Kakaoaromas aus seinen Vorstufen. Fett.

Wiss. Technol. 73(8), 515–521.Morton, I.D., Akroyd, P. and May, C.G. (1960) Flavouring substances and their preparation. GB Patent

836694.Mottram, D.S., Ames, J.M., Mlotkicwiez, J.A., Copsey, J.P. and Anderson, A. (1998) Flavouring agents.

World Patent 98/42208.Mottram, D.S. and Madruga, M.S. (1994) Important sulfur-containing aroma volatiles in meat. In: Sulfur

Compounds in Foods (eds C.J. Mussinan and M.E. Keelan), American Chemical Society, Washington,DC, pp. 180–187.

Mottram, D.S. and Nobrega, C.C. (1998) Formation of volatile sulfur compounds in reaction mixturescontaining cysteine and three different ribose compounds. In: Food Flavors: Formation, Analysis andPackaging Influences (eds E.T. Contis, C.-T. Ho and C.J. Mussinan), Elsevier Science, Amsterdam, pp.483–492.

Munch, P. and Schieberle, P. (1998) Quantitative studies on the formation of key odorants in thermally-treatedyeast extracts using stable isotope dilution assays. J. Agric. Food Chem. 46, 4695–4701.

Nedvidek, W., Ledl, F. and Fischer, P. (1992) Detection of 5-hydroxymethyl-2-methyl-3(2H)-furanone andof �-dicarbonyl compounds in reaction mixtures of hexoses and pentoses with different amines. Z.Lebens.-Unters. Forsch. 194, 222–228.

Nestle (1966) Flavouring agent. GB Patent 1032334.Ottinger, H., Bareth, A. and Hofmann, T. (2001b) Characterization of natural ‘cooling’ compounds formed

from glucose and L-proline in dark malt by application of taste dilution analysis. J. Agric. Food Chem.49, 1336–1344.

Ottinger, H., Hofmann, T. (2003) Identification of the taste enhancer alapyridaine in beef broth and evaluationof its sensory impact by taste reconstitution experiments. J. Agric. Food Chem. 51, 6791–6796.

Ottinger, H., Soldo, T., Hofmann, T. (2001a) Systematic studies on structure and physiological activity ofcyclic �-keto enamines, a novel class of ‘cooling’ compounds. J. Agric. Food Chem. 49, 5383–5390.

Parliment, T.H., McGorrin, R.J. and Ho, C.-T. (eds) (1989) Thermal Generation of Aromas, AmericanChemical Society, Washington, DC.

Parliment, T.H., Morello, M.J. and McGorrin, R.J. (eds) (1994) Thermally Generated Flavors: Maillard,Microwave, and Extrusion Processes, American Chemical Society, Washington, DC.

Pippen, E.L. and Mecchi, E.P. (1969) Hydrogen sulfide, a direct and potentially indirect contributor to cookedchicken aroma. J. Food Sci. 34, 443–446.

Pittet, A.O., Rittersbacher, P. and Muralidhara, R. (1970) Flavor properties of compounds related to maltoland isomaltol. J Agric. Food Chem. 18, 929–932.



Pittet, A.O. and Seitz, E.W. (1974) Flavoring compositions and processes. US Patent 3782973.Reineccius, G.A. (1990) The influence of Maillard reactions on the sensory properties of foods. In: The Mail-

lard Reaction in Food Processing, Human Nutrition and Physiology (eds P.A. Finot, H.U. Aeschbacher,R.F. Hurrell and R. Liardon), Birkhauser Verlag, Basel, Switzerland, pp. 157–170.

Reineccius, G.A. (1998) Kinetics of flavor formation during Maillard browning. In: Flavor Chemistry: ThirtyYears of Progress (eds R. Teranishi, E.L. Wick and I. Hornstein), Kluwer Academic/Plenum Publishers,New York, pp. 345–352.

Roberts, D.D. and Acree, T.E. (1994) Gas chromatography-olfactometry of glucose-proline Maillard re-action products. In: Thermally Generated Flavors: Maillard, Microwave, and Extrusion Processes(eds. T.H. Parliment, M.J. Morello and R.J. McGorrin), American Chemical Society, Washington, DC,pp. 71–79.

Rodel, W., Habisch, D. and Ruttloff, H. (1988) Formation of cocoa flavour by Maillard reaction. Charact.Prod. Appl. Food Flavours 2, 301–309.

Rolli, K., Roschli, D. and Sihver, J.J. (1988) Process for preparing a flavouring ingredient. European Patent286838.

Rosing, E.A.E. and Turksma, H. (1997) Process for the preparation of a savoury flavour. European Patent784936.

Rusoff, I.I. (1958) Flavor. US Patent 2835592.Rychlik, M. and Grosch, W. (1996) Identification and quantification of potent odorants formed by toasting

of wheat bread. Food Sci. Technol. (London) 29, 515–525.Schieberle, P. (1990) The role of free amino acids present in yeast as precursors of the odorants 2-acetyl-

1-pyrroline and 2-acetyltetrahydropyrridine in wheat bread crust. Z. Lebensm.-Unters. Forsch. 191,206–209.

Schieberle, P. (1991a) Primary odorants in popcorn. J. Agric. Food Chem. 39, 1141–1144.Schieberle, P. (1991b) Primary odorants of pale lager beer. Differences to other beers and changes during

storage. Z. Lebensm.-Unters. Forsch. 193, 558–565.Schieberle, P. (1992a) Bildung wichtiger Rostaromastoffe in Lebensmitteln aus Getreide. Getreide, Mehl und

Brot 46, 338–342.Schieberle, P. (1992b) Formation of Furaneol in heat-processed foods. In: Flavour Precursors – Thermal and

Enzymatic Conversion (eds R. Teranishi, G.R. Takeoka and M. Guntert), ACS Symposium Series 490,Washington, DC, pp. 164–174.

Schieberle, P. (1993a) Studies on the flavour of roasted white sesame seeds. In: Progress in Flavour Pre-cursor Studies – Analysis, Generation and Biotechnology (eds P. Schreier and P. Winterhalter), AlluredPublishing, Carol Stream, IL, pp. 343–360.

Schieberle, P. (1993b) Untersuchungen zum Aromabeitrag und zur Bildung von 4-Hydroxy-2,5-dimethyl-3(2H)-furanon in thermisch behandelten Lebensmittel. Lebensmittelchemie 47, 15–16.

Schieberle, P. (1995) Quantification of important roast-smelling odorants in popcorn by stable isotopedilution assays and model studies on flavor formation during popping. J. Agric. Food Chem. 43, 2442–2448.

Schieberle, P. (1996) Odor-active compounds in moderately roasted sesame. Food Chem. 55, 145–152.Schieberle, P. (2005) The carbon module labeling (CAMOLA) technique: a useful tool for identifying

transient intermediates in the formation of Maillard-type target molecules. Ann. N. Y. Acad. Sci. 1043,236–248.

Schieberle, P., Fischer, R. and Hofmann T. (2003) The carbohydrate module labeling technique: a usefultool to clarify formation pathways of aroma compounds formed in Maillard-type reactions. In: FlavourResearch at the Dawn of the Twenty-First Century, Proceedings of the 10th Weurman Flavour ResearchSymposium (eds. J.-L. Le Quere and P.X. Etievant), Lavoisier, Intercept, London, 447–452.

Schieberle, P. and Grosch, W. (1987) Evaluation of the flavor of wheat and rye bread crusts by aroma extractdilution analysis. Z. Lebensm.-Unters. Forsch. 185, 111–113.

Schieberle, P. and Grosch, W. (1994) Potent odorants of rye bread crust. Differences from the crumb andwheat bread crust. Z. Lebensm.-Unters. Forsch. 198, 292–296.

Schieberle, P. and Hofmann, T. (1998) Characterization of key odorants in dry-heated cysteine-carbohydratemixtures: comparison with aqueous reaction systems. In: Flavor Analysis (eds C.J. Mussinan and M.J.Morello), American Chemical Society, Washington, DC, pp. 320–330.

Schieberle, P. and Hofmann, T. (2002) New results on the formation of important Maillard aroma compounds.In: Advances in Flavours and Fragrances: From the Sensation to the Synthesis (ed. K.A. Swift), RoyalSociety of Chemistry, Cambridge, pp. 163–177.



Schieberle, P., Hofmann, T. and Munch, P. (2000) Studies on potent aroma compounds generated in Maillard-type reactions using the odor-activity-value concept. ACS Symp. Ser. 756, 133–150.

Schlichtherle-Cerny, H., Affolter, M., Blank, I., Cerny, C., Robert, F., Beksan, E., Hofmann, T. and Schieberle,P. (2002) Amadori and Heyns rearrangement products as flavoring compounds for imparting umami tasteto food products. Eur. Pat. Appl. EP 1252825 A1.

Schnermann, P. and Schieberle, P. (1997) Evaluation of key odorants in milk chocolate and cocoa mass byaroma extract dilution analysis. J. Agric. Food Chem. 45, 867–872.

Schonberg, A. and Moubacher, R. (1952) The Strecker degradation of a-amino acids. Chem. Rev. 50, 261–277.Semmelroch, P. and Grosch, W. (1995) Analysis of roasted coffee powders and brews by gas

chromatography–olfactometry of headspace samples. Food Sci. Technol. (London) 28, 310–313.Semmelroch, P., Laskawy, G., Blank, I. and Grosch, W. (1995) Determination of potent odorants in roasted

coffee by stable isotope dilution assays. Flavour Frag. J. 10, 1–7.Shibamoto, T. (1983) Heterocyclic compounds in browning and browning/nitrite model systems. In: Instru-

mental Analysis of Foods, Vol. 1 (eds G. Charalambous and G. Inglett), Academic Press, New York, pp.229–277.

Shu, C.-K. and Ho, C.-T. (1989) Parameter effects on the thermal reaction of cystine and 2,5-dimethyl-4-hydroxy-3(2H)-furanone. In: Thermal Generation of Aromas (eds T.H. Parliment, R.J. McGorrin andC.-T. Ho), American Chemical Society, Washington, DC, pp. 229–241.

Soldo, T., Blank, I., Hofmann, T. (2003) (+)-(S)-Alapyridaine – a general taste enhancer? Chem. Senses 28,371–379.

Stahl, H.D. and Parliment, T.H. (1994) Formation of Maillard products in the proline-glucose model system –high-temperature short-time kinetics. In: Thermally Generated Flavors, Maillard, Microwave and Ex-trusion Processes (eds T.H. Parliment, M.J. Morello and R.J. Mc Gorrin), American Chemical Society,Washington, DC, pp. 251–262.

Tai, C.-Y. and Ho, C.-T. (1997) Influence of cysteine oxidation on thermal formation of Maillard aromas.J. Agric. Food Chem. 45, 3586–3589.

Tandy, J.S. (1985) Chicken flavorants and the processes for preparing them. GB Patent 2157538.Teranishi, R., Buttery, R.G. and Guadagni, D.G. (1975) In: Geruchs- und Geschmacks-stoffe (ed. F. Drawert),

Verlag Hans Carl, Nurnberg, Germany, pp. 177–186.Theron, P.P.A., von Fintel, R.G.H., Saisselin, A.L. and Vissers, A.M. (1975) Flavouring agents. GB Patent

1382335.Tressl, R. (1989) Formation of flavor compounds in roasted coffee. In: Thermal Generation of Aromas (eds

T.H. Parliment, R.J. McGorrin and C.T. Ho), American Chemical Society, Washington, DC, pp. 285–301.Tressl, R., Helak, B., Kersten, E. and Rewicki, D. (1993) Formation of flavor compounds by Maillard reaction.

In: Recent Developments in Flavor and Fragrance Chemistry (eds R. Hopp and K. Mori), Verlag Chemie,Weinheim, Germany, pp. 167–181.

Tressl, R., Nittka, C. and Kersten, E. (1995) Formation of isoleucine-specific Maillard products from [1-13C]-d-glucose and [1-13C]-d-fructose. J. Agric. Food Chem. 43, 1163–1169.

Tressl, R. and Rewicki, D. (1999) Heat generated flavors and precursors. In: Flavor Chemistry: Thirty Yearsof Progress (eds R. Teranishi, E.L. Wick and I. Hornstein), Kluwer Academic/Plenum Publishers, NewYork, pp. 305–325.

Turksma, H. (1993) Process for the preparation of savoury flavours. European Patent 571031.Ullrich, F. and Grosch, W. (1987) Identification of the most intense volatile flavor compounds formed during

autoxidation of linoleic acid. Z. Lebensm.-Unters. Forsch. 184, 277–282.Van den Ouweland, G.A.M., Demole, E.P. and Enggist, P. (1989) Process meat flavor development and the

Maillard reaction. In: Thermal Generation of Aromas (eds T.H. Parliment, M.J. Morello, R.J. McGorrinand C.T. Ho), American Chemical Society, Washington, DC, pp. 433–441.

Van den Ouweland, G.A.M. and Peer, H.G. (1968) Mercapto furane and mercapto thiophene derivatives. GBPatent 1283912.

Van den Ouweland, G.A.M. and Peer, H.G. (1970) Synthesis of 3,5-dihydroxy-2-methyl-5,6-dihydropyran-4-one from aldohexoses and secondary amine salts. Recl. Trav. Chim. Pay. B. 89, 750–754.

Van den Ouweland, G.A.M. and Peer, H.G. (1975) Components contributing to beef flavor. Volatile com-pounds produced by the reaction of 4-hydroxy-5-methyl-3(2H)-furanone and its thio analog with hydrogensulfide. J. Agric. Food Chem. 23, 501–505.

Van Pottelsberghe de la Potterie, P.J. (1972) Verfahren zur Herstellung von Geschmacksstoffen mit Rinder-bratenaroma. German Patent 2149682.

Van Pottelsberghe de la Potterie, P.J. (1973) Beef flavor. US Patent 3716380.



Vauthey, S., Leser, M. and Milo, C. (1998) An aroma product comprising saturated C16 and C18 monoglyc-erides. European Patent 1008305.

Wagner, R. and Grosch, W. (1997) Evaluation of potent odorants of French fries. Food Sci. Technol. (London)30, 164–169.

Weenen, H. (1998) Reactive intermediates and carbohydrate fragmentation in Maillard chemistry. FoodChem. 62, 393–401.

Weenen, H. and Apeldoorn, W. (1996) Carbohydrate cleavage in the Maillard reaction. In: Flavour Science:Recent Developments (eds A.J. Taylor and D.S. Mottram), The Royal Society of Chemistry, Cambridge,pp. 211–216.

Weenen, H., Kerler, J. and van der Ven, J. (1997) The Maillard reaction in flavour formation. In: Flavoursand Fragrances (ed. K.A.D. Swift), The Royal Society of Chemistry, Cambridge, pp. 153–171.

Weenen, H. and Tjan, S.B. (1992) Analysis, structure, and reactivity of 3-deoxyglucosone. In: FlavorPrecursors: Thermal and Enzymatic Conversions (eds R. Teranishi, G.R. Takeoka and M. Guntert),American Chemical Society, Washington, DC, pp. 217–231.

Weenen, H. and Tjan, S.B. (1994) 3-Deoxyglucosone as flavour precursor. In: Trends in Flavour Research(eds H. Maarse and D.G. van der Heij), Elsevier Science, Amsterdam, pp. 327–337.

Weenen, H. and van der Ven, J.G.M. (1999) Formation of Strecker aldehydes. In: Book of Abstracts, 218thASC National Meeting, New Orleans, American Chemical Society, Washington, D.C.

Weenen, H., van der Ven, J.G.M., van der Linde, L.M., van Duynhoven, J. and Groenewegen, A. (1998) C4, C5,and C6 3-deoxyosones: structures and reactivity. In: The Maillard Reaction in Foods and Medicine (edsJ. O’Brien, H.E. Nursten, M.J.C. Crabbe and J.M. Ames), The Royal Society of Chemistry, Cambridge,pp. 57–64.

Werkhoff, P., Bretschneider, W., Emberger, R., Guntert, M., Hopp, R. and Kopsel, M. (1991) Recent de-velopments in the sulfur flavour chemistry of yeast extracts. Chem. Mikrobiol. Technol. Lebensm. 13,30–57.

Whitfield, F.B. (1992) Volatiles from interactions of Maillard reactions and lipids. Crit. Rev. Food Sci. Nutr.31, 1–58.

Whitfield, F.B. and Mottram, D.S. (1999) Investigation of the reaction between 4-hydroxy-5-methyl-3(2H)-furanone and cysteine or hydrogen sulfide at pH 4.5. J. Agric. Food Chem. 47, 1626–1634.

Widder, S., Sabater Luntzel, C., Dittner, T. and Pickenhagen, W. (2000) 3-Mercapto-2-methylpentan-1-ol, anew powerful aroma compound. J. Agric. Food Chem. 48, 418–423.

Yaylayan, V.A., Forage, N.G. and Mandeville, S. (1994) Microwave and thermally induced Maillard reactions.In: Thermally Generated Flavors: Maillard, Microwave, and Extrusion Processes (eds T.H. Parliament,M.J. Morello and R.J. McGorrin), American Chemical Society, Washington, DC, pp. 449–456.

Yaylayan, V.A. and Huyghues-Despointes, A. (1994) Chemistry of Amadori rearrangement products: anal-ysis, synthesis, kinetics, reactions and spectroscopic properties. Crit. Rev. Food Sci. Nutr. 34, 321–369.

Yaylayan, V.A. and Keyhani, A. (1999) Origin of 2,3-pentanedione and 2,3-butanedione in d-glucose/l-alanine Maillard model systems. J. Agric. Food Chem. 47, 3280–3284.

Yaylayan, V.A. and Keyhani, A. (2000) Origin of carbohydrate degradation products in l-alanine/d-[13C]glucose model systems. J. Agric. Food Chem. 48, 2415–2419.

Yaylayan, V. and Sporns, P. (1987) Novel mechanisms for the decomposition of 1-(amino acid)-1-deoxy-d-fructoses (Amadori compounds): a mass spectrometric approach. Food Chem. 26, 283–305.

Yeretzian, C., Blank, I. and Palzer, S. (2007) Process flavourings. In: Flavourings. Production, Composition,Applications, Regulations (ed. H. Ziegler), Wiley-VCH, Weinheim, Germany, pp. 549–572.

Zamora, R. and Hidalgo, F.J. (2005) Coordinate contribution of lipid oxidation and Maillard reaction to thenonenzymatic food browning. Crit. Rev. Food Sci. Nutr. 45, 49–59.

Zhang, Y. and Ho, C.-T. (1989) Volatile compounds formed from thermal interaction of 2,4-decadienal withcysteine and glutathione. J. Agric. Food Chem. 37, 1016–1020.

Zhang, Y. and Ho, C.-T. (1991) Formation of meatlike aroma compounds from thermal reaction of inosine5′-monophosphate with cysteine and glutathione. J. Agric. Food Chem. 39, 1145–1148.

Zheng, Y., Brown, S., Ledig, W.O., Mussinan, C. and Ho, C.-T. (1997) Formation of sulfur-containing flavorcompounds from reactions of Furaneol and cysteine, glutathione, hydrogen sulfide, and alanine/hydrogensulfide. J. Agric. Food Chem. 45, 894–897.

3 basic chemistry and process conditions for reaction ... · 3 basic chemistry and process...

Documents