Food and Packaging Interactions II - American Chemical Society


Food and Packaging Interactions II - American Chemical Societyhttps://pubs.acs.org/doi/pdf/10.1021/bk-1991-0473.ch012Sim...

0 downloads 66 Views 1MB Size

Chapter 12 Thermodynamics of Permeation of Flavors in Polymers Prediction of Solubility Coefficients 1

2

3

G. Strandburg , P. T. DeLassus , and B. A. Howell

Downloaded by GRIFFITH UNIV on October 26, 2017 | http://pubs.acs.org Publication Date: September 13, 1991 | doi: 10.1021/bk-1991-0473.ch012

1

3

Center for Applications in Polymer Science, and Department of Chemistry, Central Michigan University, Mt Pleasant, MI 48859 Barrier Resins and Fabrication Laboratory, The Dow Chemical Company, 1603 Building, Midland, MI 48674

2

The use of polymers as food packaging materials has experienced rapid and continuous growth in recent years. This trend is likely to endure (1). Polymer packages have evolved from simple food wraps to sophisticated containers which have additional demands placed upon them. The package must participate in the flavor management of the food. Flavor degeneration is a complex issue, and various modes of loss can occur. Of primary importance is oxidation resultingfrominvasion of oxygenfromthe atmosphere through the container and subsequent reaction with the food. Elimination of this type of flavor degeneration requires that the package restrict the transport of oxygen. It must also limit the flux of water vapor. It is critical that moist foods do not dehydrate and that dry foods remain dry. Flavor degeneration can also resultfrommigration of large molecules, typically organic solvents and flavor and aroma molecules. These will contain four to twelve carbon atoms and may contain functional groups as well. Contamination of the foodfroman external source is one mode of flavor alteration. An example of this process is migration of perfumes used in detergents and health and beauty aids into food packages during storage. Ingress of solvents used in printing inks and adhesives are also representative of this method of flavor degeneration. An area of greater concern is the migration of flavor and aroma moleculesfromthe food. Flavor and aroma molecules are present in extremely small quantities; often the total concentration is less than one part per million. However these molecules are responsible for the unique flavor of a particular food, and small losses often result in dramatic off taste. Losses can be from broadremovalof all flavors or an imbalance caused by selectiveremovalof only a few flavor components. Flavor losses thatresultfromthe interaction with the polymer package can be classified into two categories. First, losses occurring by permeation or migration through the package, and, second,thosefromsorption or scalping by the container. Many plastic packages contain a barrier polymer to minimize flavor losses. Barrier 0097-6156/91/0473-0133S06.00/0 © 1991 American Chemical Society

Risch and Hotchkiss; Food and Packaging Interactions II ACS Symposium Series; American Chemical Society: Washington, DC, 1991.

Downloaded by GRIFFITH UNIV on October 26, 2017 | http://pubs.acs.org Publication Date: September 13, 1991 | doi: 10.1021/bk-1991-0473.ch012

134

FOOD AND PACKAGING INTERACTIONS

polymers are polymers that inhibit mass transport. Suppression of mass transport must be non-selective; it must be effective for both large and small molecules. The purpose of this study was to investigate methods to quantify these losses in typical packaging polymers. More important, given the growth of plastics as food packaging materials coupled with the complexity of flavor make-up, it would be advantageous to be able to predict mass transport parameters of flavor molecules in commercial polymers. Included in this work are semi-empirical methods to predict mass transport parameters for a variety of molecules in selected polymers. The permeation process can be described as a multi-step event First, collision of the penetrant molecule with the polymer is followed by sorption into the polymer. Next, migration through the polymer matrix by random hops occurs, and finally, desorption of the permeant from the polymer completes the process. The process occurs to eliminate an existing chemical potential difference. Steadystate mass transport in polymers with a constant chemical potential follows Fick's first law of diffusion. For a polymer having a thin film geometry, Fick's first law can be written as, AM

X

= PAAp At L

(1)

x

where A M / A t is the transport rate of material x through a film of area A , having a thickness of L , and under a chemical potential created by pressure difference across the film of A p . P is the permeability coefficient and is a steady-state parameter. In this study S.I. units will be used, thus the permeability coefficient will be reported as kg*m/m *s*Pa. Equation 2 is used to describe the mass transport process. X

x

2

P=

D*S

(2)

The permeability as discussed earlier is a steady-state parameter. It consists of two component parts, the diffusion coefficient (D) and the solubility coefficient (S). The diffusion coefficient is a kinetic parameter. It is a measure of how fast transport events will occur. It reflects the ease with which a penetrant molecule moves within a polymer host. More specifically, the time required to reach steadystate transport is provided by this parameter. It provides an estimate of the effective depth of penetration of a permeant into the matrix as a function of time. Knowledge of the diffusion coefficient is crucially important in applications where steady state is not reached. The solubility coefficient is a thermodynamic parameter. It is a measure of the concentration of penetrant molecules that will be in position to migrate through the polymer. The solubility coefficient is an equilibrium partition coefficient for distribution of the penetrant between polymer and vapor phase such that the following equation holds.

c = s* x

Px

Risch and Hotchkiss; Food and Packaging Interactions II ACS Symposium Series; American Chemical Society: Washington, DC, 1991.

(3)

12. STRANDBURG ET AL.

135

Permeation of Flavors in Polymers

Where C is the concentration in the polymer and p is the partial pressure of x in the vapor phase. S is analogous to a reciprocal Henry's law constant. The solubility coefficient is dependent on many variables. Of primary importance is the condensability of the penetrant. The solubility coefficient is also a function of the thermodynamic interaction of the penetrant with the polymer. Losses of flavor/aroma molecules into the polymer package can be quantified from a knowledge of the diffusion and solubility coefficients. While numerous models exist to estimate the diffusion process in polymers, little pertaining to the thermodynamic parameter has been reported. This study was undertaken to introduce fundamental understanding of the thermodynamics involved in the mass transport process. Arguments throughout will be supported with experimental data. Downloaded by GRIFFITH UNIV on October 26, 2017 | http://pubs.acs.org Publication Date: September 13, 1991 | doi: 10.1021/bk-1991-0473.ch012

x

x

EXPERIMENTAL A variety of methods exist for measuring permeability, diffusion, and solubility coefficients for large molecules. These methods have been described elsewhere (2). The derivative method was used exclusively in this study. The polymers were studied as mono-layer structures having thin film geometry. Figure 1 represents detector response as a function of time for this technique. The detector response in the derivative method reflects the mass transport rate AM /At. At tQ a film, free of penetrant molecules, is exposed to penetrant vapor on one side, hereafter the upstream side. The downstream side is continually swept by a carrier gas to a detector. As a result the downstream side of the film is always at zero partial pressure of penetrant. Initially the mass transport rate is below the detector limits. As time progresses the transport rate increases to allow detection. The transport rate rises steadily through a transient region and eventually levels off at a steadystate rate at which time AM^/At is constant. After calibration, the permeability coefficient can be obtained from the detector response using equation 1. The diffusion coefficient was determined from equation 4 (3). X

D

=

L

2

7.2*t

(4) 1 / 2

s

In this expression D is the diffusion coefficient, L is the film thickness and t ^ * the time required to reach one half the steady-state mass transport rate. Once the permeability coefficient and the diffusion coefficient have been obtained, the solubility coefficient can be calculated using equation 2. Figure 2 is a schematic of the instrument used for this study. It has been described in detail elsewhere (4). The instrument consists of a gas handling system integrated with a mass spectrometer. The gas handling system contains the plumbing, glassware, mixing pumps, permeation cell, and switching valves necessary for a permeation experiment. The entire system is housed in an insulated gas chromatograph air bath oven. Temperature control is from sub-ambient to 150"C±1*C.

Risch and Hotchkiss; Food and Packaging Interactions II ACS Symposium Series; American Chemical Society: Washington, DC, 1991.

Downloaded by GRIFFITH UNIV on October 26, 2017 | http://pubs.acs.org Publication Date: September 13, 1991 | doi: 10.1021/bk-1991-0473.ch012

136

FOOD AND PACKAGING INTERACTIONS

Figure 1. Typical Detector Response Curve for the Derivative Technique for the Determination of Mass Transport Parameters.

HP 5970

Mass Spectrometer

Control & Analysis Computer

— » ^ CRT^ Display

Gas Handling

Figure 2. Schematic Diagram for the Permeation Instrument

Risch and Hotchkiss; Food and Packaging Interactions II ACS Symposium Series; American Chemical Society: Washington, DC, 1991.

Downloaded by GRIFFITH UNIV on October 26, 2017 | http://pubs.acs.org Publication Date: September 13, 1991 | doi: 10.1021/bk-1991-0473.ch012

12. STRANDBURG ET AL.

Permeation of Flavors in Polymers

137

The mass spectrometer employed is a Hewlett Packard Model 5970B Mass Selective Detector. The mass spectrometer can be operated in scan mode or in a selected-ion monitoring mode. Data acquisition, storage, and editing are accomplished with a Hewlett Packard 59970C Chem Station. The mass spectrometer was used in two ways. First, mass spectra of permeants and solvents were generated. Ion fragments from the spectra were then selected for the permeation experiments. The criterion for selection was based on the most populous ion fragments that were not degenerate with those from the dilution solvent used in the calibration standard. After selection of the ions, the mass spectrometer is programmed for the selected ion monitoring (SIM) mode. In SIM the mass spectrometer monitors only ions that have been selected. Typically three ion fragments were monitored in each experiment Subsequent to collecting permeation data all flow paths and flasks were checked to be free of penetrant. This was accomplished by sequentially setting the switching valves to each position required in an experiment while monitoring the visual display. If substantial amounts of permeant were detected, the experiment was aborted and the gas handling system was purged with nitrogen carrier gas until free of any contaminants. After the gas handling system was deemed to be free of permeant, the experiment was started. The first portion of the experiment was for calibrating the mass spectrometer. Initially, a baseline was established for the calibration process. After the baseline was obtained, a known quantity, typically 2-3 id, of the calibration standard (0.5% vol/vol dilution) was added to the one-liter flask. This quantity generated a partial pressure of penetrant of about 0.3 Pa (three parts-per-million mole/mole) in nitrogen varying somewhat with the formula weight and density of the penetrant. Following a sufficient mixing time, the calibration standard was sent to the mass spectrometer. After the detector signal for the calibration standard leveled off, a known quantity, usually 2-3 /zl, of neat permeant was added to the three-liter flask and was allowed to mix. The partial pressure in the three-liter flask was about 20 Pa (200 parts-per-million mole/mole) of permeant in nitrogen. After a constant calibration response was obtained, the standard was exhausted and a baseline for the experiment was established. At a recorded time, tQ, the penetrant vapor was allowed to circulate on one side of the film while the opposite side was continually swept to the mass spectrometer. The experimental progression was followed by visual observation of the monitor. Following the experiment the gas handling system was purged with carrier gas to remove excess permeant Data from the experiment were retrieved, and conversion to a hard copy for analysis completed the experimental process. Figure 3 is a total ion chromatogram for the transport of ethyl valerate through a vinylidene chloride copolymer film at 105" C. The initial curve represents the calibration portion of the experiment and is followed by the permeation portion. The experimental design for this study had three parts. First, a rigorous set of experiments to measure the mass transport coefficients for a complete homologous family of compounds through one polymer film was conducted. Many

Risch and Hotchkiss; Food and Packaging Interactions II ACS Symposium Series; American Chemical Society: Washington, DC, 1991.

138

FOOD AND PACKAGING INTERACTIONS

of these experiments were done at small temperature intervals, and multiple experiments were performed at selected temperatures. The data base was used to identify and quantify tendencies as the penetrant was changed in a controlled manner while the polymer was held constant. The goal of this experimentation was to establish the ability to predict the mass transport characteristics for other compounds within this family. The permeants used in this study were linear esters from Flavor and Fragrances Kit No. 1 from The Aldrich Chemical Company. These compounds were used without further purification. The esters used appear in Table I along with selected physical properties.

Downloaded by GRIFFITH UNIV on October 26, 2017 | http://pubs.acs.org Publication Date: September 13, 1991 | doi: 10.1021/bk-1991-0473.ch012

Table I. Linear Esters for Permeation Studies Ester name

Methyl butyrate Ethyl propionate Ethyl butyrate Propyl butyrate Ethyl valerate Ethyl hexanoate Ethyl heptanoate Ethyl octanoate Hexyl butyrate

Boiling Point CC)

102 99 120 142 144 168 188 206 208

Purity (%)

98+ 97+ 98+ 98+ 98+ 98+ 98+ 98+ 98+

CAS Number

623-42-7 105-37-3 105-54-4 105-66-8 539-82-2 123-66-0 106-30-9 106-32-1 2639-63-6

The polymer chosen for the first part of the experimental design was a copolymer of vinylidene chloride and vinyl chloride henceforth co-VDC. It is a commercial polymer used as a barrier layer in food packaging applications. The coV D C has extremely small diffusion coefficients and thus low steady-state mass transport rates. Elevated temperatures were required (85" C -105 * C) for more timely results.

The second part of the experimental process was to obtain mass transport coefficients for the ester family in another polymer. In this portion of the process a subset of the esters was used. The esters included methyl butyrate, propyl butyrate, ethyl hexanoate, and ethyl heptanoate. The polymer used in this section was low density polyethylene (LDPE). L D P E is not a barrier for the transport of these compounds but represents an important packaging material. Another reason for including L D P E was that mass transport data already exist for selected molecules through this material. Hence, mass transport measurements of properly chosen esters through LDPE allow for a check of the experimental technique.

Risch and Hotchkiss; Food and Packaging Interactions II ACS Symposium Series; American Chemical Society: Washington, DC, 1991.

12. STRANDBURG ET AL.

Permeation of Flavors in Polymers

139

The third part of this study was an attempt to expand the general observations of the mass transport coefficients for the esters to analogous processes for other homologous families. The compounds chosen to complete this section of the study were linear ketones and n-alkanes. These compounds were chosen since the size and shapes are similar to those of the linear esters. Therefore differences observed for these compounds would be expected to arise from functional group differences. The ketones and alkanes were obtained from The Aldrich Chemical Company and used without further purification. Selected members from these families and some characteristic parameters appear in Table n . Comparing results

Downloaded by GRIFFITH UNIV on October 26, 2017 | http://pubs.acs.org Publication Date: September 13, 1991 | doi: 10.1021/bk-1991-0473.ch012

from the three families should provide insight with respect to mass transport for many different polymer/penetrant systems. Table II. Ketones and n-Alkanes Used in Permeation Studies Compound Name

Boiling Point CC)