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

Подождите немного. Документ загружается.

of the frequency window of dynamic measurements.

Plotting the data as J

(w), J

(w) is sometimes more

telling, as shown in Figure 14 for the same gluten as in

Figure 13; J

(w) displays a maximum at the right of

the frequency range, which at least allows location of

the upper end of the viscoelastic plateau on the time

scale and possibly opens the way to modeling this

region of the behavior.

0038Enlarging the time scale of observation is required

anyway for a full-scale rheological characterization.

In the case of classical polymer systems, the time–

temperature superposition principle, which means

that a change in the temperature of the system is

equivalent to a shift of the time scale of the behavior,

allows one to describe the viscoelastic behavior over a

very large frequency range from data obtained over

−4

−3

−2

−1

G',G" (Pa)

ω (rad s

−1

)

fig00 13 Figure 13 Mechanicalspectraobtainedat20



Cbydynamicmeasurementsonvarioussystemsofthecolloidalgeltype.Circles:

fullyhydratedglutenfromacommercialwheatcultivar.Triangles:heat-setgelofbovineserumalbumin(61mgml

1

,pH7,0.05moll

1

NaCl).Squares:experimentalyogurt.Lines:commercialmayonnaise.Filledsymbolsandsolidline:G

;emptysymbolsandinterrupted

line:G

−2

−3

−4

−5

−3

−2

−1

J',J" (m

−1

)

ω (rad s

−1

)

fig00 14 Figure 14 Mechanicalspectrum ofthesam eglutenasshownin Figure 13,butplottedasJ

andJ

versusangularfrequency.The

curve of J

shows a peak with a maximum close to o ¼ 0.16 rad s

1

; it bounds the viscoelastic plateau on the high-frequency side.

4976 RHEOLOGICAL PROPERTIES OF FOOD MATERIALS

a restricted one at different temperatures. This

principle does not in general apply to food systems

since their elasticity is rarely entropy-driven; anyway,

the range of temperature which could be covered is

quite narrow because of the presence of liquid water.

One has then to combine dynamic measurements

with transient tests which permit one to extend the

observation window to longer times. Let us give a few

illustrations.

0039A gluten and a reconstituted dough are compared

in Figure 15. The mechanical spectra of a gluten and

of a ‘synthetic’ dough prepared by mixing it with

starch do not differ much on the 10

1

–10

rad s

1

range (Figure 15a). The moduli are somewhat larger

for dough, because starch granules act as a filler;

in fact, their presence also makes dough behavior

extremely nonlinear and the dynamic data are only

approximate. But when the materials are submitted

−3

(a)

−2

−1

G',G" (Pa)

G' gluten

G" gluten

G' dough

G" dough

ω (rad s

−1

)

−4

−3

−2

(b) 10

t (s)

J(t) (m

−1

)

fig0015 Figure 15 Viscoelastic behavior at 20



C of a reconstituted dough compared to that of the gluten used to prepare it. (a) Mechanical

spectra obtained by dynamic measurements under 0.05 strain amplitude over the 10

2

–10

rad s

1

frequency range. (b) Creep curves

under 350 Pa stress. Symbols: creep compliance J(t); lines: recoverable compliance J

(t). Filled symbols and dotted line, gluten; empty

symbols and solid line, dough.

RHEOLOGICAL PROPERTIES OF FOOD MATERIALS 4977

to a retardation test in the same conditions, the

difference becomes striking after 100-s creep (Figure

15b); analysis of the results shows that whereas J



about the same (* 10

3

1

), viscosity is much

smaller for dough (* 510

Pa s

1

) than for gluten

(10

Pa s

1

) at comparable shear rates (7 10

4

versus 2 10

4

1

0040 When the tests are performed within the linearity

range, it is possible to convert the data of relaxation

or retardation tests from the time to the frequency

domain. Combining the data of the retardation test

with those of dynamic measurements allowed us to

obtain the mechanical spectrum of a weak gluten

sample over about eight logarithmic frequency

decades (Figure 16). The viscoelastic plateau extends

over more than seven frequency logarithmic decades,

the terminal region of the spectrum being entered

only below *10

5

rad s

1

. Extended viscoelastic

plateaus result from strong interactions and multi-

level structures within the material. Such cases are

not exceptional in food systems and highlight the

fact that rheology is a matter of time.

See also: Colloids and Emulsions; Emulsifiers: Organic

Emulsifiers; Phosphates as Meat Emulsion Stabilizers;

Uses in Processed Foods; Gums: Properties of Individual

Gums; Nutritional Role of Guar Gum; Pectin: Properties

and Determination; Food Use; Protein: Sources of Food-

grade Protein; Quality; Heat Treatment for Food Proteins;

Starch: Structure, Properties, and Determination;

Sources and Processing; Functional Properties.

Further Reading

Axelos MAV, Renard D, Chevillard C and Lefebvre J (1999)

Viscoelastic properties of some biopolymer systems in

relation with their nano-structure. In: Stokke BT and

Elgsaeter A (eds) Wiley Polymer Group Review Series,

vol. 2, pp. 330–342. New York: Wiley.

Barnes HA (2000) A Handbook of Elementary Rheology.

Aberystwyth: Institute of Non-Newtonian Fluid

Mechanics.

Barnes HA, Hutton JF and Walters K (1989) An Introduc-

tion to Rheology. Amsterdam: Elsevier.

Ferry JD (1980) Viscoelastic Properties of Polymers, 3rd

edn. New York: Wiley.

Hill SE, Ledward DA and Mitchell JR (1998) Functional

Properties of Food Macromolecules, 2nd edn. Gaithers-

burg: Aspen.

Macosko CW (ed) (1994) Rheology. Principles, Measure-

ments, and Applications. New York: VCH.

Marin G (1988) Oscillatory rheometry. In: Collyer AA

and Clegg DW (eds) Rheological Measurement, pp.

297–343. London: Elsevier.

Powell RL (1988) Rotational viscometry. In: Collyer AA,

Clegg DW (eds) Rheological Measurement, pp.

247–296. London: Elsevier.

Rao MA and Steffe JF (eds) (1992) Viscoelastic Properties

of Foods. London: Elsevier.

Ross-Murphy SB (1994) Rheological methods. In: Ross--

Murphy SB (ed) Physical Techniques for the Study of

Food Biopolymers, pp. 343–392. London: Blackie.

Whorlow RW (1992) Rheological Techniques, 2nd edn.

Chichester: Ellis Horwood.

−6

−5

−4

−3

−2

ω (rad s

−1

)

G',G" (Pa)

−1

fig0016 Figure 16 An example of a combination of the retardation test with dynamic measurements in order to describe the viscoelastic

behavior of a material over a large time scale. The material is a fully hydrated gluten from an experimental wheat line. Mesurements

were performed within the linearity range at 20



C. The data from the recovery part of the retardation test were converted from the time

to the frequency domain using an approximation method. Symbols: dynamic measurements: filled symbols, G

; empty symbols, G

Lines: recovery data: solid line, G

, dotted line, G

4978 RHEOLOGICAL PROPERTIES OF FOOD MATERIALS

RHEOLOGY OF LIQUIDS

A E Bell, University of Reading, Whiteknights,

Reading, UK

This article is reproduced from Encyclopaedia of Food Science,

Background

0001 Food materials are multiphase, multicomponent

systems whose rheological characteristics frequently

present process engineers with difficult handling and

pumping problems.

0002 The food industry has, almost since its incep-

tion, been concerned with the physical properties of

its products, in terms of both their processing and

their eating characteristics. Consequently, consider-

able care must be taken in designing plant and process

if the texture, consistency, and quality of the final

product are not to be compromised. The selection

of appropriate pumps and conveying and mixing

systems is especially germane to food materials that

may be shear-dependent and/or abrasive in nature.

(See Rheological Properties of Food Materials.)

0003 Although it is a relatively easy matter to describe

mathematically the flow behavior of ‘simple’ liquids,

food solid/liquid mixtures present unique problems in

describing their flow behavior.

Liquid Behavior

0004 When assessing the behavior of a liquid system, it is

usual to investigate the stresses generated when the

sample is sheared, that is to say, the force per unit area

produced when the ‘layers’ of liquid move over each

other in the direction of flow. This gives rise to the

classical definition of a shear viscosity (Figure 1):

0005 Shear rate, g ¼V/h (s

1

)

0006 Shear stress, s ¼F/A (Pa)

0007 Shear viscosity, Z ¼s/g (Pa s),

where V is the velocity, F is the force, and A is the area

over which it acts.

0008 The range of shear rates observed in the food indus-

tries is vast, from values as low as 10

2

1

, for the

self-leveling of a liquid due to surface tension, to

1

seen in the spray drying of slurries. (The

chewing and swallowing of foods generate shear

values of about 10

–10

1

.) This presents us with

considerable difficulties with regard to most food

systems, as very few behave in an ideal or Newtonian

manner, with the shear stress proportional to the

shear rate over all of the possible shear rate values.

0009Since viscosity is essentially related to shear rate

and stress, it is possible to operate a viscometer in

one or two modes: (1) by applying a rotation and

measuring the torque generated (controlled stain),

or (2) by applying a fixed torque and measuring the

speed generated (controlled stress). These two

methods have their advantages and disadvantages;

for example, the controlled-strain instruments are

relatively cheap, robust, and well understood but

usually have a limited shear range (1–1000 s

1

). Con-

trolled-stress instruments, however, have large shear

rate ranges and can carry out other forms of testing

(creep, oscillatory, etc.) but do tend to be somewhat

expensive.

Working Liquids

0010When we work a liquid system, we apply energy to it.

This ‘force’ has a number of characteristics that

we can define: (1) the area over which it is applied,

(2) the distance over which it acts, and (3) the rate at

which it is applied. We may then define the stresses

involved (force/area) and the strains produced (rela-

tive deformation). If a more complex analysis is re-

quired, these forces may be resolved further into three

components, the tensile or normal forces, perpendicu-

lar to the plane of the applied force, and two shear or

tangential forces parallel to it.

0011The strain (distance moved) when processing or

measuring a sample has a considerable effect on its

mechanical behavior/flow properties. If a sample is

moved such that its molecular domains or structural

elements no longer interact/overlap, the network may

become ‘overloaded’ and ultimately break down

(Figure 2). For example, a sample of an emulsion

V.F.

fig0001Figure 1 Classical shear viscosity. Reproduced from Rheology

of Liquids, Encyclopaedia of Food Science, Food Technology and Nu-

trition, Macrae R, Robinson RK and Sadler MJ (eds), 1993, Aca-

demic Press.

RHEOLOGY OF LIQUIDS 4979

such as mayonnaise, which is stabilized by electro-

static or steric means, may have a very low value of

‘critical strain,’ after which structural damage to the

emulsion will occur, and it is very important that

testing and processing be carried out in such a

way so as not to exceed this point. (See Colloids and

Emulsions.)

0012 While, in general, increasing the strain tends to

reduce interactions, increasing the shear rate may

give rise to either increases or decreases in the appar-

ent degrees of interaction. If we consider the effect of

increasing the shear rate on entangled long-chain

polymer systems (e.g., a long-chain polysaccharide),

at zero shear rate, these chains will be entangled by

Brownian motion and will have a degree of inter-

action based on physical constraints. On moving

one of the chains slowly with respect to the others,

very little resistance would be encountered, as the

molecules would have sufficient time to allow any

rearrangements needed to take place. However, if

the same process were carried out at a higher rate,

the entanglement would have insufficient time for

rearrangement and would resist such movement.

This would produce a material that appeared to

increase its degree of interaction at higher shear

rates. Such a system may become unworkable if, for

example, one attempted to pump it too quickly

(Figure 3). (See Carbohydrates: Classification and

Properties.)

0013 However, if a system were stabilized by, for

example, some form of hydrogen bonding, an in-

crease in shear rate might reduce the effectiveness of

the interactions. This would lead to a fall in the

apparent consistency (Figure 4).

0014 The two parameters, strain and rate, when com-

bined give an overall picture of the behavior of the

liquid. This rheological ‘plane’ may be very flat (basic

Newtonian behavior) or, more commonly, steeply

curving (nonideal behavior). The consequences of

not having assessed the physical characteristics over

a suitable testing regime may prove disastrous in

terms of predicting processing behavior.

Behavior as a Function of Time

0015In the description of the different types of non-

Newtonian behavior, it was implied that, although

the viscosity of a fluid might vary with shear rate, it

was independent of the length of time that the shear

rate was applied, and also that replicate determin-

ations at the same shear rate would always produce

the same viscosity. This must be considered as the

ideal situation, since most non-Newtonian food ma-

terials are colloidal in nature, and as such, the flowing

Rate

'Consistency'

fig0003Figure 3 Increasing interaction with shear rate. Reproduced

from Rheology of Liquids, Encyclopaedia of Food Science, Food

Technology and Nutrition, Macrae R, Robinson RK and Sadler MJ

(eds), 1993, Academic Press.

Strain

'Consistency'

fig0002 Figure 2 Effect of increasing strain. Reproduced from Rhe-

ology of Liquids, Encyclopaedia of Food Science, Food Technology

and Nutrition, Macrae R, Robinson RK and Sadler MJ (eds),

1993, Academic Press.

Rate

'Consistency'

fig0004Figure 4 Decreasing interaction with shear rate. Reproduced

from Rheology of Liquids, Encyclopaedia of Food Science, Food

Technology and Nutrition, Macrae R, Robinson RK and Sadler MJ

(eds), 1993, Academic Press.

4980 RHEOLOGY OF LIQUIDS

elements, whether particles or macromolecules, may

not adapt immediately to the new conditions. There-

fore, when such a material is subjected to a particular

shear rate, the shear stress and, consequently, the

viscosity will decrease with time. Furthermore, once

the shear stress has been removed, even if the struc-

ture that has been broken down is reversible, it may

not return to its original structure (rheological ground

state) instantly. The common feature of all these

materials is that, if they are subjected to a gradually

increasing shear rate followed immediately by a shear

rate decreasing to zero, the downcurve will be dis-

placed with regard to the upcurve, and the rheogram

will exhibit a hysteresis loop.

0016 In the case of plastic and pseudoplastic materials,

the downcurve will be displaced to below the upcurve

(Figure 5), whereas for dilatant substances, the re-

verse will be true (Figure 6).

0017The presence of the hysteresis loop indicates that a

‘breakdown’ in structure has occurred, and the area

with the loop may be used as an index of the degree of

‘breakdown.’

0018The term that is used to describe one such type of

behavior is thixotropy, which means ‘to change by

touch.’ Strictly, this term should be applied only to an

isothermal sol–gel transformation. It has, however,

become common to describe as thixotropic any

material that exhibits a reversible time-dependent

decrease in apparent viscosity. Thixotropic systems

are usually composed of asymmetric particles or

macromolecules that are capable of interacting by

numerous secondary bonds to produce a loose three-

dimensional structure, so that the material is gel-like

when unsheared. The energy that is imparted during

shearing disrupts these bonds so that the flowing

elements become aligned, and the viscosity falls,

since a gel–sol transformation has occurred. When

the shear stress is eventually removed, the structure

will tend to reform, although the process is not im-

mediate and will increase with time as the molecules

return to the original state under the influence of

Brownian motion. Furthermore, the time taken for

recovery, which can vary from minutes to days

depending upon the system, will be directly related

to the length of time that the material was subjected to

the shear stress, since this will affect the degree of

breakdown.

0019In some cases, the structure that has been destroyed

is never recovered, no matter how long the system is

left. Repeat determinations of the flow curve will then

only produce the downcurve that was obtained in the

experiment that resulted in the material’s destruction.

Such behavior should be referred to as ‘shear destruc-

tion’ rather than thixotropy, which, as will be appre-

ciated from above, is a misnomer.

0020An example of such behavior are the gels produced

by high-molecular-weight polysaccharides, which are

stabilized by large numbers of secondary bonds. Such

systems undergo extensive reorganization during

shearing such that the three-dimensional structure is

reduced: the gel-like nature of the original is then

never recovered.

0021The occurrence of such complex behavior creates

problems in quantitative classification because not

only will the apparent viscosity change with shear

rate, but there will be two ‘viscosities’ that can be

calculated for any given shear rate (i.e., from the

upcurve and the downcurve). It is usual to attempt

to calculate one viscosity for the upcurve and another

for the downcurve. This must assume, of course, that

each of the curves achieves linearity over some of its

length; otherwise, a defined shear rate must be used:

only the former situation is truly satisfactory. Each of

Rate

Stress

fig0006 Figure 6 Dilatant behavior. Reproduced from Rheology of

Liquids, Encyclopaedia of Food Science, Food Technology and Nutri-

tion, Macrae R, Robinson RK and Sadler MJ (eds), 1993, Aca-

demic Press.

Rate

Stress

fig0005 Figure 5 Plastic/pseudoplastic behavior. Reproduced from

Rheology of Liquids, Encyclopaedia of Food Science, Food Technol-

ogy and Nutrition, Macrae R, Robinson RK and Sadler MJ (eds),

1993, Academic Press.

RHEOLOGY OF LIQUIDS 4981

the lines used to derive the viscosity may be extrapo-

lated to the shear stress axis to give an associated

yield value. However, only that derived from the

upcurve has any significance, since that derived

from the downcurve will relate to the broken-down

system. Consequently, the most useful index of thix-

otropy can be obtained by integration of the area

contained within the loop. Of course, this does not

take into account the shape of the up- and down-

curves, and consequently, two materials may produce

loops of similar area but have completely different

shapes representing totally different flow behaviors.

In order to prevent confusion, it is best to adopt a

method whereby an estimate of area is accompanied

by yield value(s). This is of particular importance

with flow curves that exhibit complex upcurves.

0022 This situation is typical of the type of loop obtained

with, for example, some samples of white soft paraf-

fin, where the upcurve exhibits a number of bulges

(Figure 7). Those at a lower shear rate are thought

to be associated with the initial loss of the three-

dimensional structure, while the smoother deviations

occurring at the higher shear rates are associated with

molecular reorientation. Such behavior is common in

food and pharmaceutical systems and is one of the

major causes of difficulties in their evaluation.

0023 With such a wide range of rheological behavior, it

is extremely important to carry out measurements

that will produce meaningful results. It is crucial,

therefore, not to use a determination of viscosity at

one shear rate (such as would be acceptable for a

Newtonian fluid), since it could lead to completely

erroneous comparative results.

0024Figure 8 shows rheograms that are an example of

four different types of flow behavior, all of which

intersect at point A, which is equivalent to a shear

rate of 100 s

1

. Therefore, if a measurement were

made at this one shear rate, all four materials would

be shown to have the same viscosity, although they all

possess different properties and behavior. Single-point

determinations are probably an extreme example but

are used to emphasize the importance of properly

designed experiments.

0025It should be noted that while more complex rheo-

logical measurements may give rise to parameters

other than viscosity (for example, the storage and

loss moduli), the arguments given here still apply,

and time and shear dependence remain major prob-

lems when measuring changing food systems.

See also: Carbohydrates: Classification and Properties;

Colloids and Emulsions; Rheological Properties of

Food Materials

Further Reading

Ferry JD (1980) Viscoelastic Properties of Polymers, 3rd

edn. New York: Wiley.

Larson RG (1998) The Structure and Rheology of Complex

Fluids, 688 pp. Oxford, UK: Oxford Press.

Lewis MJ (1987) The Physical Properties of Foods and Food

Processing Systems. Chichester, UK: Ellis Horwood.

Rate

Stress

fig0008Figure 8 Convergence of flow behavior at a single rate. Repro-

duced from Rheology of Liquids, Encyclopaedia of Food Science,

Food Technology and Nutrition, Macrae R, Robinson RK and Sadler

MJ (eds), 1993, Academic Press.

Rate

Stress

fig0007 Figure 7 Complex flow behavior. Reproduced from Rheology

of Liquids, Encyclopaedia of Food Science, Food Technology and Nu-

trition, Macrae R, Robinson RK and Sadler MJ (eds), 1993, Aca-

demic Press.

4982 RHEOLOGY OF LIQUIDS

RIBOFLAVIN

Contents

Properties and Determination

Physiology

Properties and Determination

R Bitsch, Friedrich-Schiller-University, Jena,

Germany

Physicochemical Characteristics

0001 Riboflavin, the name prescribed for vitamin B

the International Union of Pure and Applied Chemis-

try – International Union of Biochemistry, is the bio-

logically active component of the prosthetic group of

flavoproteins. Older designations are lactoflavin,

ovo-, or uroflavin, which originated from the first

isolation. It comprises a tricyclic, nitrogen-containing

ring-system, chemically defined as a substituted iso-

alloxazine with a ribitol side-chain at N

. With the

systematical nomenclature, the molecule is named

7,8-dimethyl-10-(1-d-ribityl)-benzo[g]pteridin-2,4-

dion. The empiric formula is given as C

with a molecular weight of 376.36 (Figure 1). The

substance is a yellow to orange–yellow powder and

crystallizes in sharp, optically active needles, melting

under decomposition at 275–282



C. The optical ro-

tation is ½a

¼112 to  122



C at a concentration

of 50 mg in 10 ml of 0.02 N alcoholic NaOH). Ribo-

flavin is sparingly soluble in water with a yellow–

green fluorescence, very sparingly soluble in ethanol,

and insoluble in other organic solvents. It is readily

soluble in dilute alkali, but rapidly deteriorated, being

accelerated by light. In neutral or acid solutions, it is

more stable, even against heat and atmospheric

oxygen. Nevertheless, all operations with riboflavin

solutions should be handled in the dark or in red light.

The light sensitivity of the flavins also has conse-

quences for the stability of vitamin B

containing

foods. In milk stored in day light, the riboflavin

content decreases by more than 80% within hours.

Processing losses in foods are given with approxi-

mately 20%.

HCOH

O ~ O

Ribit ~ Adenin

Riboflavin

FMN (H

)

FAD (H

)

OH OH

+2[H]

−2[H]

fig0001 Figure 1 Riboflavin and its coenzyme forms.

RIBOFLAVIN/Properties and Determination 4983

0002 Light exposure of alkaline solutions decomposes

riboflavin to lumiflavin (7.8.10-trimethylisoalloxa-

zine), and in neutral or acidic solutions, the blue

fluorescent lumichrome (7.8-dimelthylalloxazine) is

formed (Figure 2 and Table 1). Riboflavin is found

in free form only in the retina of the eye, whey, and

urine. In other biological systems and tissues, the

predominant forms are flavin mononucleotide

(FMN) or riboflavinphosphate and flavin adenine

dinucleotide (FAD). Free riboflavin and FMN-mono-

sodium salt are the most important commercial prod-

ucts. FMN and FAD are effective as coenzymes or

prosthetic groups in more than 100 enzymes in the

plant and animal kingdoms. The accepted designa-

tion for both compounds is incorrect, because FMN is

not truly a nucleotide and FAD is not a dinucleotide.

The designations arose from the discovery of ribofla-

vin, and have been adopted ever since. Pure riboflavin

was first isolated from yeast, egg white, and whey

in 1933 (by Kuhn and Wagner-Jauregg). Simultan-

eously, the so-called ‘yellow coenzymes’ were

discovered (by Warburg and Christian) in yeast and

could be identified as FMN. FMN is considerably

more soluble in water than riboflavin. Its solubility

in organic solvents and its stability against heat and

light exposure are comparable with that of riboflavin.

The flavoproteins FMN and FAD belong to the en-

zymatic class of oxidoreductases. The conjugated

structure of the pteridine moiety of the basic molecule

determines its coloring and light absorbance in the

UV and visible spectrum. Absorbance maxima for the

flavins are listed in Table 1. Likewise, catalytic hydro-

gen transfer in the flavin enzymes takes place at both

conjugated of bound N

and N

atoms of the isoal-

loxazin molecule. The oxidized form of the flavopro-

teins, the flavoquinone, is thereby reversibly reduced

to the colorless flavohydroquinone or leucobase (see

Figure 1). The absorbance is lost, because double

bonds are dissolved. As an intermediate product

during this reaction, a radical semiquinone is formed.

During the redox process, the flavin molecule changes

its conformation, which is planar in the oxidized state

Lumiflavin

(alkaline solution)

Lumichrome

(acidic solution)

fig0002 Figure 2 Photolysis products of riboflavin (lumiflavin and lumichrome).

tbl0001 Table 1 Physicochemical data of flavines

Property Riboflavin Riboflavin- 5 -phosphate-Na (FMN-Na) Flavin adenine dinuclectide-Na

(FAD-Na

)

Molecular weight 376.36 478.34 829.6

Melting point 278–282



C (decomp.) 280–290



C (decomp.)

Solubility Readily soluble in dilute alkalis

(decomp.)

Soluble in water at pH 7

(30 g l

1

,25



Soluble in water

Sparingly soluble in water

(70–100 mg l

1

,27



Very sparingly soluble in ethanol

(45 mg l

1

,27



Very sparingly soluble in ethanol

Insoluble in acetone, chloroform, ether, benzene

Absorbance (UV/vis) 223, 267, 374, 444 nm (0.1 N–HCI)

Fluorescence Yellow–green fluorescence of aqueous solutions at pH 6–7

(ex. 444 nm, em. 530–565 nm)

ex. 400–500 nm, em. 530 nm

Stability In neutral and acidic medium relatively stable against heat and oxygen. Rapid photolysis by UV- and visible

light to blue fluorescent lumichrome (acidic and neutral solutions) or yellow–green fluorescent lumiflavin

(alkaline solutions). After reduction by dithionite, Zn–HCI, sodium hydrosulfite, etc. reversible conversion

to leucoflavin (dihydroflavin)

According to The Merck Index, 11th edn. Rahway NJ: Merck & Co. and modified from Eitenmiller RR and Landen WO, Jr. (1999) Vitamin Analysis for the

Health and Food Sciences, pp. 229–337. Boca Raton, FL: CRC Press.

4984 RIBOFLAVIN/Properties and Determination

and folded in a butterfly configuration in the reduced

state.

0003 The characteristic yellow–green fluorescence of

riboflavin and FMN in aqueous solutions exhibits

maximal intensity at pH 6–7. The fluorescence of

FAD is about 10–20% of that of riboflavin with the

same spectral distribution but exhibiting a maximal

intensity at pH 3. The fluorescence of flavins is

quenched when bound to proteins. The irradiation

products of the flavins also fluoresce, and their spec-

trum is shifted to the blue (lumichrome) or green

(lumiflavin) spectral range. Riboflavin can be quanti-

fied by conversion to lumiflavin and by measuring its

fluorescence intensity in chloroformic solutions with

exitation (ex.) at 270 nm, and emission (em.) at

418 nm.

Methods of Determination

0004 Initially, direct fluorimetry of the intact flavins or

after hydrolysis to FMN or free riboflavin have been

most frequently used for quantitation of vitamin B

active components. Alternatively, the lumiflavin

method based upon the conversion of the chloro-

form-insoluble riboflavin into chloroform-soluble

lumiflavin after photodegradation in alkaline solu-

tion can be used, despite a transformation rate of

only 60–70%. Currently, microbiological assays and

HPLC procedures with an enhanced sensitivity are

widely used (see also Table 2).

0005 In principle, riboflavin is extracted from dry mater-

ials by autoclaving in 0.1 N HCl for 30 min or in a

water bath for 1–2 h. By using acid extraction, flavins

are released from their protein binding, and FAD and

FMN are converted to free riboflavin. In protein-rich

materials, acidic extraction may be followed by en-

zymatic hydrolysis for complete liberation of flavins.

It is important to carry out all operations relating to

the extraction and quantitation of flavins and ribofla-

vin under dim light and in brown or dark vessels.

Direct Fluorimetry

0006Direct measurement of the native fluorescence of

riboflavin may be applicable for pharmaceutical

preparations or food samples with a relatively high

vitamin content. After acid extraction, any remaining

proteins are removed by precipitation at pH 4.5.

Interfering substances are subsequently oxidized

with potassium permanganate and excess oxidant

destroyed with hydrogen peroxide or sodium bisul-

fite. The fluorescence of the clear sample solution,

commonly spiked with riboflavin as an internal

standard, is registered by difference measurement

before and after reduction of riboflavin with sodium

bisulfite to the leuco form for correction by unspecific

fluorescence (ex. ¼440 nm; em. ¼565 nm).

0007The detection limit of this method is about 0.1 mg

of riboflavin per gram of sample. The direct fluori-

metry of flavins had been adopted by the US Pharma-

copoeia for the riboflavin analysis in tablets and

injections, and by the AOAC for the riboflavin deter-

mination in foods, as in milk-based formulas, for

example.

Lumiflavin Method

0008In this method, riboflavin is photodegraded in alka-

line solution at pH 10–12 to the stronger fluorescent

lumiflavin (ex. ¼270 nm; em. ¼418 nm). This reac-

tion is specific for riboflavin, since other flavins inter-

fering with riboflavin fluorescence are not converted

into lumiflavin. Lumiflavin is separated from ribofla-

vin and other fluorescent substances by extraction

with chloroform. The transformation into lumiflavin

is not quantitative and ranges between 60 and 70%.

Reliable results can be obtained only by comparison

tbl0002 Table 2 Methods of riboflavin determination

Method Principles Specials Detectionlimits

Direct fluorimetry Fluorescence measurement of riboflavin at

440 (ex.)/565 (em.) nm

Conversion of FAD and

FMN to riboflavin by

acidic hydrolysis

0.1 mgB

per gram of sample

Lumiflavin method Fluorescence measurement after irradiation

at pH 10–12 (ex. 270, em. 418 nm)

Lumiflavin is separated by

chloroform extraction

0.02 ng of B

Microbiological assay Turbidimetric measurement of bacterial

growth response dependent on the

content

Interference of growth

response with nutrients/

food components

0.1–0.5 ng of B

HPLC analysis Separation of flavins on RP-C

materials

with aqueous/organic solvents and

fluorimetric/UV detection

Differentiated analysis of

flavins, combined

analysis with other

B-vitamins

0.21 ng of riboflavin

0.89 ng of FMN

11.15 ng of FAD

0.02 ng of B

(precolumn

irradiation to lumiflavin)

RIBOFLAVIN/Properties and Determination 4985