Patent Application
20190248566
The present application relates to methods and systems for optimizing plastic compositions formed from resin feedstocks, and more specifically to optimal blending or the use of modifiers/additives in recycled thermoplastic in packaging to increase shelf-life of perishable products.
In recent years, federal and state legislation has been proposed to increase the use of sustainable (bio-based) and recycled packaging materials. The adoption of recycled plastic or bio-plastic alternatives has been slow due to contaminant loads, the costs associated with collection and sorting, and overall quality. Bio-plastics have been viewed as a costly alternative to traditional plastics. Thus, there is a need to incentivize the use of bioplastics and recycled packaging materials.
Agricultural growers and packers utilize plastic packaging for retail convenience foods and food service products and are currently the largest users of plastic film and sheeting in the United States. Studies have indicated that the vitamin content of current fruits and vegetables is significantly lower than 50 years ago. This is important as most Americans do not eat enough of these products.
It has been suggested that the increased haze, scattering, and filtering of light in recycled and bio-plastic substrates reduces the rate of product degradation. As such, recent initiatives by retailers have called for an increase in the use of recycled packaging substrates. However, the nature of recycled packaging material, which includes variance in the overall composition, makes it difficult to predict how effective the final plastic product will be at providing these benefits. Further, once the product is finalized, there is no opportunity to adjust the composition to provide the maximum benefit with respect to increasing the shelf-life of perishable products.
The present invention is directed to overcoming these and other deficiencies in the art.
One aspect of the present invention relates to a method of optimizing a plastic composition formed from a plurality of resin feedstocks. A plurality of resin feedstocks are provided. The plurality of resin feedstocks are blended to form the plastic composition. One or more properties of the plastic composition, including radiation absorption, radiation transmission, gas evolution, fluorescence, or melting properties, are measured. The ratio of the plurality of resin feedstocks being blended into the plastic composition is adjusted, based on the measured one or more properties, to form an optimized plastic composition.
Another aspect of the present invention relates to a system including a blending apparatus configured to blend a plurality of resin feedstocks. One or more sensors are positioned at different locations in the blending apparatus to measure one or more properties of the plurality of resin feedstocks, including radiation absorption, radiation transmission, photoionization, or melting properties. The system further includes a computing device comprising a processor and a memory coupled to the processor which is configured to execute one or more programmed instructions stored in the memory and comprising receiving measurements of the one or more properties of the plurality of resin feedstocks from the one or more sensors and outputting one or more instructions to the blending apparatus to blend the plurality of resin feedstocks based on the measured one or more properties to optimize the plastic composition. A compound delivery system is configured to blend the plurality of resin feedstocks based on the one or more properties.
This technology provides increases in nutritive retention and shelf-life, as well as perceived freshness of vegetables, when utilizing recycled and bio-plastics of varying composition. Many compounds have been identified as contributing to the light filtering effect seen in recycled plastics. The present technology allows for the presence of these compounds to be modified by optimal blending or by using additives s part of the recycling process or during conversion and manufacturing to optimize the resultant plastic composition to provide extended shelf-life for products stored therein.
Utilizing the present technology, various radiation, absorption, emission, fluorescence, and thermal properties of plastic compositions are measured during the conversion of resin feedstocks to a plastic composition. The measured properties provide data that may be utilized to adjust the blending of the resin feedstocks or to add additional materials to the blend to optimize the resultant plastic composition to provide longer shelf-life and reduce degradation for products stored in the plastic composition.
The present application relates to methods and devices for optimizing plastic compositions formed from resin feedstocks, and more specifically to the use of modifiers in recycled thermoplastic in packaging to increase shelf-life of perishable products.
One aspect of the present invention relates to a system including a blending apparatus configured to blend a plurality of resin feedstocks. One or more sensors are positioned at different locations in the blending apparatus to measure one or more properties of the plurality of resin feedstocks, including radiation absorption, radiation transmission, photoionization, or melting properties. The system further includes a computing device comprising a processor and a memory coupled to the processor which is configured to execute one or more programmed instructions stored in the memory and comprising receiving measurements of the one or more properties of the plurality of resin feedstocks from the one or more sensors and outputting one or more instructions to the blending apparatus to blend the plurality of resin feedstocks based on the measured one or more properties to optimize the plastic composition. A compound delivery system is configured to blend the plurality of resin feedstocks based on the one or more measured properties.
An exemplary environment 10 including a computing device 12 coupled to one or more sensors 14(1)-14(n) and a compound delivery system 15 by a communication network 18 is illustrated in
Referring again more specifically to
Processor 20 in computing device 12 executes a program of instructions stored in the memory for one or more aspects of the present technology, although other numbers and types of systems, devices, components, and elements in other configurations and locations can be used.
Memory 22 in computing device 12 stores the programmed instructions for one or more aspects of the present technology, although some or all of the programmed instructions could be stored and/or executed elsewhere. A variety of different types of memory storage devices, such as a random access memory (RAM), read only memory (ROM), hard disk, CD ROM, DVD ROM, or other computer readable medium which is read from and written to by a magnetic, optical, or other reading and writing system that is coupled to processor 20, can be used for memory 22.
Communication interface 24 of the computing device 12 is used to operatively couple and communicate between the computing device 12 and sensors 14(1)-14(n) via the communications network 18, although other types and numbers of communication networks, systems, or other links with other types and numbers of connections and configurations can be used. By way of example only, communications network 18 could use TCP/IP over Ethernet and industry-standard protocols, including NFS, CIFS, SOAP, XML, LDAP, and SNMP, although other types and numbers of communication networks, such as a direct connection, a local area network, a wide area network, modems and phone lines, e-mail, and wireless communication technology, each having their own communications protocols, can be used.
Input device 26 and display device 28 of computing device 12 enable a user to interact with computing device 12, such as to input and/or view data and/or to configure, program, and/or operate computing device 12. Input device 26 may include a keyboard, computer mouse, and/or touch screen and display device 28 may include a computer monitor, although other types and numbers of input devices and/or display devices could also be used in other examples. Input device 26, by way of example, may be utilized to manually input data regarding one or more properties of the plastic composition.
Sensors 14(1)-14(n) may be any sensors known in the art capable of measuring one or more properties of a plastic composition in accordance with the examples of methods illustrated and described herein. Although a plurality of sensors 14(1)-14(n) are illustrated, it is to be understood that a single sensor could be utilized in some examples. By way of example, sensors 14(1)-14(n) are configured to measure radiation absorption, radiation transmission, gas evolution, radiation fluorescence, or melting properties of a plastic composition.
Sensors 14(1)-14(n) may be configured to perform, by way of example, ultraviolet-visible spectroscopy analysis, an attenuated total reflectance Fourier transform infrared spectroscopy analysis, a differential scanning calorimetry analysis, a mechanical analysis, x-ray fluorescence analysis, or energy dispersive x-ray fluorescence analysis of the plastic composition. It is to be understood that sensors 14(1)-14(n) are not limited to these measurement techniques, and other analytical techniques may be employed that are suitable for obtaining one or more properties of a plastic composition.
In one example, an ultraviolet visible spectrometer may be utilized. The ultraviolet visible spectrometer must have sufficient resolution to generate a spectrum of absorbance as a function of the wavelength of the incident irradiation between 300 nm and 400 nm. The range of absorption measurement analysis may be lower than 300 nm and higher than 400 nm depending on the desired absorption properties of the product. The thickness of the specimen analyzed (perpendicular to the incident irradiation) is measured as the thickness of the specimen affects the measured absorption properties. A single wavelength absorption measurement may be sufficient if the data represents the desired ultraviolet/visible absorption properties or under rapid throughput conditions for quality control.
In another example, Fourier transform infrared spectroscopy is utilized. A background spectrum is collected and stored. The minimum number of scans collected for the background spectrum will be considered as the number of scans that generates a representative spectrum of the background environment. The collected spectra may be normalized (i.e., the spectrum is multiplied by a number) to the intensity of a wavenumber that is representative of a characteristic band of the polymer that does not change in intensity due to a chemical reaction. In addition, transmission experiments may be normalized by the path length of the specimen (typically the thickness of the specimen). In certain cases, a mathematical subtraction may be performed to emphasize spectral shifts according to equation 1.
A(v)subtracted spectrum=A(v)spectrum A−A(v)spectrum B [1]
In yet another example, fluorescence intensity measurements are made with an instrument capable of irradiating the sample with ultraviolet and visible light and detecting the fluorescence response. The instrument must at minimum be able to irradiate the sample at 350 nm excitation wavelength, however, other excitation wavelengths may be used for analysis. The fluorescence detector must at minimum be able to detect the intensity of the fluorescence emission at 501 nm, however, the fluorescence intensity of other emission wavelengths may be used for analysis so long as the data are indicative of the properties indicated in this invention. In some cases, a “3D” scan can be employed where the fluorescence intensity of at a given wavelength for a specific excitation wavelength for multiple sequential excitation wavelengths at a known interval.
Sensors 14(1)-14(n) are located at positions 102-109 in extrusion system 16 as illustrated in
Compound delivery system 15 is coupled to extrusion system 16 and is configured to blend the plurality of resin feedstocks in the extrusion system 15 based on the one or more properties of the plastic composition measured by sensors 14(1)-14(n) as illustrated in
Although an example of environment 10 including computing device 12 is described herein, it is to be understood that the devices and systems of the examples described herein are for exemplary purposes, as many variations of the specific hardware and software used to implement the examples are possible, as will be appreciated by those skilled in the relevant art(s).
Aspects of the examples may also be embodied as a non-transitory computer readable medium having instructions stored thereon for one or more aspects of the present technology as described and illustrated by way of the examples herein, as described herein, which when executed by a processor, cause the processor to carry out the steps necessary to implement the methods of the examples, as described and illustrated herein.
Another aspect of the present invention relates to a method of optimizing a plastic composition formed from a plurality of resin feedstocks. A plurality of resin feedstocks are provided. The plurality of resin feedstocks are blended to form the plastic composition. One or more properties of the plastic composition, including radiation absorption, radiation transmission, gas evolution, radiation fluorescence, or melting properties, are measured. The ratio of the plurality of resin feedstocks being blended into the plastic composition is adjusted, based on the measured one or more properties, to form an optimized plastic composition.
An exemplary method for optimizing a plastic composition in accordance with the present technology will now be described with respect of
In step 300, a plurality of resin feedstocks are provided. In one example, the resin feedstocks are thermoplastic resin feedstocks, such as polyethylene terephthalate (PET), although other thermoplastic resin feedstocks such as polyethylene, polypropylene, polystyrene, poly methyl methacrylate, polycarbonate, an addition polymer, a condensation polymer, and mixtures thereof, may be utilized. The plurality of resin feedstocks may include a recycled polymeric material and a virgin polymeric material, although in one example only recycled polymeric material is utilized. The plurality of resin feedstocks may be provided in various combinations depending upon the application.
Next, in step 302, the plurality of resin feedstocks are blended to form the plastic composition. The plurality of resin feedstocks are blended in a blending apparatus, such as extrusion system 16 illustrated in
In step 304, one or more properties of the plastic composition are measured by sensors 14(1)-14(n) coupled to extrusion system 16. Although a plurality of sensors 14(1)-14(n) are illustrated, it is to be understood that a single sensor could be utilized at single measurement location. The one or more measured properties of the plastic composition include one or more of radiation absorption, radiation transmission, gas evolution, radiation fluorescence, or melting properties. In one example, the measuring in step 304 is carried out at least two different points in time following the blending in step 302. The measuring in step 304 may be performed using ultraviolet-visible spectroscopy analysis, an attenuated total reflectance Fourier transform infrared spectroscopy analysis, a differential scanning calorimetry analysis, a mechanical analysis, x-ray fluorescence analysis, or energy dispersive x-ray fluorescence analysis. It is to be understood that the measuring performed is not limited to these measurement techniques, and other analytical techniques may be employed that are suitable for obtaining one or more properties of a plastic composition. In one example, the measuring is performed by sensors 14(1)-14(n), which may be located at various locations along extrusion system 16, such as locations 102-109 illustrated in
Next, in step 306, computing device 12 receives the measured properties from sensors 14(1)-14(n) through communication network 18. Alternatively, the one or more properties may be measured by sensors 14(1)-14(n) and entered manually into computing device 12 through input device 26. Computing device 12 may receive data from sensors 14(1)-14(n) at multiple points in time during the conversion process.
In step 308, computing device 12 correlates the measured properties of the plastic composition to radiation transmission, radiation absorption, radiation emission, or radiation reflection properties for a completed plastic composition product having those measurements.
In step 310, computing device 12 identifies an adjustment to the blending of resin feedstocks or one or more additives to be added to the plastic composition to optimize these properties. More specifically, if the absorption properties desired are more similar to the absorption properties of a composition comprising 75% RPET and the measured absorption properties are more closely associated with 25% RPET, then the computing device 12 can alter the feed ratio of the resin feedstocks utilizing a calibrated single wavelength measurement (e.g., 350 nm) to adjust the composition ratio to achieve the desired absorption spectrum as described in
In step 312, the ratio of the plurality of resin feedstocks being blended into the plastic composition is adjusted, based on said measuring, to form an optimized plastic composition. The plastic composition is optimized based on radiation transmission, radiation absorption, radiation emission, or radiation reflection properties. The radiation transmission, radiation absorption, radiation emission, or radiation reflection properties are optimized so the optimized plastic composition isolates and controls electromagnetic wavelengths associated with one or more of vitamin degradation, adverse color changes, chlorophyll degradation, or degradation of other nutritional components. In one example, the plastic composition blocks between 50% to 75% more of incident ultraviolet light compared to the plastic composition not containing recycled content at an exemplary thickness of 20 microns. The optimized properties allow for increased shelf-life of products stored in the plastic composition. By way of example, the blending of the plastic composition is optimized to isolate certain destructive wavelengths for particular vitamins/compounds as illustrated in Table 1 below, which identifies wavelengths associated with nutrient degradation loss for the listed nutrients.
Table 1 below, which identifies wavelengths associated with nutrient degradation loss for the listed nutrients.
In one example, the blending is carried out using compound delivery system 15, such as a co-extruder, a dosing pump, or a direct intake to an extrusion line. Compound delivery system 15 may be located at any one of the locations 102-109 in extrusion system 16 as illustrated in
One or more additive compounds selected from the group consisting of thermal or light stabilizers, antioxidants, plasticizers, fillers, nucleating agents, colorants, thermal conductors, catalysts, and combinations thereof, are added to the plastic composition to provide the optimization in step 312. The one or more additives may be added using compound delivery system 15. Additive compounds are selected, by way of example, to increase light filtering to extend shelf-life. Utilizing this process allows manufacturers to control specifications and monitor the direct or indirect addition of organic and inorganic compounds to feedstock to allow for maximum shelf-life extension, while maintaining customer performance.
Ultraviolet radiation is broken down into three specific wavelength sections: UVA, UVB, and UVC. Generally, UVC (100-280 nm) is completely absorbed by the atmosphere and UVB (280-315 nm) has a low depth of penetration due to their short wavelengths and therefore typically only result in surface interactions. However, the wavelengths associated with UVA radiation (315-400 nm) are long enough to penetrate tissue and possess sufficient energy to yield detrimental degradation. UV-Vis spectra of the PET sheet absorbed more than the limits of the equipment between 280 and 315 nm independent of RPET concentration, and thus, are expected to perform similarly for absorption of UVB radiation. However, the absorbance of UVA radiation absorption increased with % RPET concentration (
Thermo-mechanical processing of polyethylene terephthalate is known to cause main chain degradation of the polymer producing a multitude of degradation byproducts. Thermo-oxidation of PET commonly occurs at the diethylene glycol constituent forming a peroxide compound. The resulting hydroxyl radical reacts with the benzene ring of the terephthalic acid constituent producing quinone/hydroquinone derivatives which are known to absorb UV irradiation (
The loss of carbonyl groups detected by attenuated total reflectance Fourier transform infrared spectroscopy increased as a function of increasing recycled content (
As discussed above, degradation of PET occurs as a result of thermomechanical processing. The mechanisms of degradation in the literature propose the production of quinone derivatives on the aromatic ring of the terephthalic acid constituent which are known to fluoresce. The increased fluorescence intensity between 470 and 570 nm is a result of increased concentration of the fluorescing moiety, which would be produced as a result of degradation events (
Previous work indicated that green vegetables when packaged in recycled content PET containers and exposed to fluorescent light at 276 lux, yellowed and senesced at much faster rates than when stored under dim conditions at 36 lux (
Freshly-harvested chives (Allium schoenoprasum L.) were obtained from a commercial grower in San Diego, Calif. Prior to shipment, the herbs were placed into polyethylene bags with perforations, then immediately cooled by forced air to approximately 4° C. The chives were placed in expanded polystyrene-lined cardboard boxes with pre-cooled gel bags, and shipped via overnight mail. Upon arrival, the herbs were held at 1±0.5 C, and 90% RH. The herbs were used within twenty-four hours.
The chives were placed into non-vented, 907 ml produce containers which were subsequently covered with shrink-wrap film with moderate barrier properties. Each packaging treatment was replicated 4 times with 3 samples per replicate. A weight of 60±2 g of chives was placed into each container and sealed. Packages were placed into a constantly illuminated (fluorescence) controlled environment room held at 1±0.5 C. 90% RH. Samples were randomly placed within the chamber. Average light intensity varied from 2760 lux at the top shelves to 750 lux at the middle shelves, to 340 lux at the bottom shelves of the chambers.
Weights for packages were recorded at the time they were placed into storage. At one-week intervals, two randomly-chosen samples from each treatment were removed and weighed. Samples were compared with fresh, loose samples shipped from the supplier. Variables were rated hedonically on a scale of 1-9, with 1 representing best quality and 9, complete breakdown and tissue destruction. Each package was evaluated for odor, color, wilt, and decay. Values were combined and averaged to produce an overall quality rating. However, when the value for any one variable exceeded 5, the product was considered to be no longer marketable.
Samples containing known amounts of recycled content can extend shelf-life up to 14 days (340 Lux vs 2760 Lux) utilizing filtering mechanisms and through recycled content formulation (
Manufacture and Testing of Post-Consumer Recycled Sheet:
Recycled PET (RPET) of varying compositions and compounds were dried/recrystallized (<0.2%) using a Farragtech Card E series dryer/crystallizer (Farrag Tech GmbH; Wolfurt, Austria) and blended at approximately 0, 20, 40, 60, 80, and 100% virgin-to-recycled content with varying compound loads as illustrated in Table 2 below. Each converted sheet using RPET blends with compositions of 0, 20, 40, 60, 80, and 100% of virgin/recycled resin was extruded on a single-screw extruder (Davis-Standard, Pawcatuck, Conn.) to a final thickness of 17-20 mil. Sheets were then formed into lidded punnets using a straight vacuum Formech Midi former (Formech Machine; Chicago, Ill.). The thickness of each specimen was recorded at 12 random locations using a digital micrometer with a resolution of ±0.5 μm.
Determination of Thermal Properties of Manufactured Sheet:
Differential scanning calorimetry was used to determine the thermal properties of the molded sheets (
Determination of Light Absorption Properties of Manufactured Sheet:
Ultraviolet-visible (UV-Vis) spectroscopy was used to determine the absorption properties of each PET/RPET sheet in the UV-Vis regions (
Chemical Migration of Recycled Polyethylene Terephthalate (RPET):
Domestically manufactured, thermoformed polyethylene terephthalate (PET) containers with label claims of 0, 50, 70, or 100% post-consumer (PC) recycled water bottle content were collected from retail stores. The food simulants specified by 21CFR177.1630 that were used in this study were n-heptane, 8% ethanol in water (v/v), and water. N-heptane (99% Optima grade, Fisher Scientific, Fair Lawn, N.J.) was used as received from the manufacturer. Deionized water (16.7 mega-ohm) was produced using a Barnstead Nanopure II (Dubuque, Iowa) water purification system. A 95% ethanol solution (ACS Spectrophotometric grade, Acros Chemicals, Morris Plains, N.J.) was diluted to 8% (v/v) with nanopure deionized water.
Preparation of Samples:
The experimental design was a completely randomized 3×3 factorial. A circular disk was cut from each package, slightly larger than the area of exposure to ensure a good seal between the sample and the analysis apparatus: the area of exposure to the food simulants was 42.65 cm2 for each sample. Samples were conditioned according to ASTM D618-13 (ASTM 2013) using a Thermo-Forma Scientific environmental chamber coupled to Watlow (Winona, Minn.) 982 series controllers. The mass of each disk was determined using a Mettler Toledo (Columbus, Ohio) Model AB104 scale with a resolution of +/−0.1 mg.
Chemical Migration Analyses:
Samples were analyzed for chemical migration according to Title 21, Chapter 1, Subchapter B, Part 177.1630 of the Code of Federal Regulations (CFR 2014). This code requires that each package must not desorb more than 0.5 mg of total chemicals per square inch (6.45 cm2) into specified food simulants: nanopure deionized water, 8% ethanol solution in nanopure deionized water, and n-heptane.
Gas Chromatography-Mass Spectroscopy:
After exposure to the test samples, aliquots of the solvents were injected into a Clarus 600 GC (Perkin Elmer, Shelton, Conn.) equipped with a 0.25 mm I.D., 30 m RTX 502.2 capillary column (Restek, Bellefonte, Pa.) attached to a Perkin Elmer Clarus 600C mass spectrometer (MS). The helium carrier gas was set to a constant flow rate of 2.5 mL/min with the inlet port set to 280° C. The following oven temperature ramping protocol was used: initial temperature 35° C. for 5 min, then increased to 150° C. at a rate of 30° C./min, then increased to 250° C. at a rate of 20° C./min. The final temperature of 250° C. was held for 10 min. Continuous mode MS at a rate of 5 scans/s between m/z ratios of 35 and 300 was employed. Aliquots of pure solvent were analyzed by GC-MS to produce background spectra, the peaks of which were eliminated from the spectra obtained from the food simulants in the 21CFR177.1630 analysis.
Elemental Analysis by Inductively Coupled Plasma-Atomic Emission Spectroscopy:
The PET/RPET sheets were analyzed by inductively coupled plasma-atomic emission spectroscopy (ICP-AES) for elemental composition. California Health and Safety Code 25214.13 states that lead, mercury, cadmium, and hexavalent chromium are contaminants of particular concern and are considered to be regulated metals. One specimen of each sheet was analyzed for elemental composition to confirm that introducing recycled-PET into extruded sheets for food and cosmetic packaging applications meets the requirements of the Toxics in Packaging Prevention Act (California Health and Safety Code 25214.13).
Example compliment additives associated with recycled plastic and biopolymer materials can be found below in ranges that provide optimal light, color and optical properties to increase freshness of products contained.
Example compounds and loading affecting product quality and freshness.
Compound loading to maximize performance and develop standards for recycled thermoplastics (% RPET) was determined using the “best” subsets of predictors of % RPET from a set of 17 potential indicators. The analysis proceeded in two steps: the first step identified the subset of variables that were independent of the silicone coating, a nuisance factor. From the derived subset, the second step identified the best subsets of predictors of % RPET that maximize performance and ranges.
The first task was to identify a subset of predictor variables that were independent of the nuisance factor, the silicone coating. For this purpose, the silicone variable was scored 1=coating and 0=no coating and used as the dependent variable in a binary logistic model as described, by way of example, in Cox, D. R. “The Continuity Correction” Biometrika 57:217-219 (1970) and Cox et al., Theoretical Statistics (Chapman and Hall, London) (1974), the disclosures of which are hereby incorporated herein by reference in their entirety. To accomplish this objective, the efficient score as disclosed in Rao, C. R., “Linear Statistical Inference and its Applications,” 2nd ed., New York: Wiley (1973), which is hereby incorporated herein by reference in its entirety, was obtained for the 17 indicators (Table 3).
Rao's efficient score, λ, measures the initial contribution a variable makes in predicting an outcome; it is commonly used as an initial screen in forward selection algorithms. In typical forward selection algorithms, if a potential predictor has a sufficiently significant λ-score; it becomes eligible for further analysis. However, in the current study the goal was to identify indicators that were independent of the nuisance factor; thus, non-significant linear combinations of indicators were sought. In line with this objective, the eight variables reported in Table 3 with λ-scores with p>0.10 were flagged as potentially independent of the nuisance factor, the silicone coating. To identify the linear combination of variables, independent of a silicone coating, the systematic “directed search” selection method of Daniel et al., “Fitting Equations to Data: Computer Analysis of Multifactor Data,” J. Wiley and Sons, New York, N.Y. (1980), the disclosure of which is hereby incorporated herein by reference in its entirety, was adapted as follows: Variables with the smallest λ-scores (high p values) were entered into the binary logistic regression one at a time. After each entry, a xx2 test for linear dependence and the Cox-Snell RR2 were calculated. This process stopped when a statistically significant linear combination was obtained (Table 4).
The systematic test of nested variables revealed five variables that were independent of the nuisance factor: A350 nm, Tm″ width, Tc, Tm′, and % Crystal. The linear combination independence requirement breaks down when the sixth variable, Tm″ onset, is entered in the binary logistic regression, xx2=18.86, df=6, p=0.004, and the Cox-Snell RR2=0.27 (Table 4).
Generalized linear modelling, GZLM, is a general method allowing for various model types. The current model is a generalized logit model, one with a Binomial prior and a uniform observed data distribution. The posterior is a mixture of the two distributions.
For the present data set (Yi, Xi, wi) the following definitions apply, i=1, . . . , n denotes the 60 observations, Yi is a prior binomial count corresponding to % RPET, Xi is the vector of the predictor variables (Table 2), and wi is the observed data weight for criterion Yi. The goal of the analysis was to identify the “best” subsets of % RPET predictors. For this purpose the generalized logit model for extra-binomial variation was employed.
The current experiment has six levels of Yi(% RPET): 0, 20, 40, 60, 80, 100% and to denote the number of observations, i=1, 2, . . . , n. Yi is the imaginary prior count of the number of “successes” out of mi=m=100 trials. For each of the six levels of % RPET there are 10 observations for a total of n=60 observations. Following William's (1982) model II, the expected value and variance are given as:
E(Yi)=mπi [2]
Var(Yi)=φwi−1mπi(1−πi) [3]
where φwi−1 represents the extra-binomial variation, φ is given by [l+ρ(m−1)], and where ρ is a parameter that may be interpreted as the correlation between the binary components; wi−1 is the scale weight. We assume constant ρ for all proportions. The goal of the analysis was to identify the “best” subsets of % RPET predictors. For this purpose, a generalized logit model for extra-binomial variation was employed.
The generalized logit model is defined by two functions g and v and the link function g(πi)=ηi is a monotonic differentiable link function relating the mean probability (or binomial proportion)
to the linear predictor η=Xβ where X is an n×p design matrix, β is a p×1 vector of regression coefficients. The variance function vv relates the variance with the expected binomial count, mπi, to the variance by Var(Yi)=ϕw−1v(πi) where ϕ=mφ. For the binomial distribution v(πi)=πi(1−πi) so the scale factor ϕ=1 and wi=1 under ideal binomial conditions. In this experiment, however, the observed values for Yi are uniformly distributed over six levels: 0, 20, 40 60, 80, and 100; thus, the observed data weight reflects a uniform distribution, wi=⅙.
The five potential predictors (Table 4) were studied in the following analysis. The generalized logit model was defined as the following link-linear function:
ηi=g(πi)=Σjβjxij, i=1, . . . ,n. [4]
with the logit link function g such that,
From the five eligible variables (Table 4), the next step identified the “best” subsets. The five variables give 26 subset models; each is evaluated using the Bayesian information criterion, BIC, and the finite sample corrected Akaike Information Criterion, AICc.
where L is the log likelihood function, kk=pp+1 is the number of predictors plus the intercept in the model, nn is the number of observations and mm is the number of trials. BIC and AICc were used as the information criteria to evaluate the relative quality of the generalized logit models as disclosed in Akaike, H. “Maximum Likelihood Identification of Gaussian Autoregressive Moving Average Models,” Biometrika 60(2):255-265. (1973); Schwarz, G., “Estimating the Dimension of a Model.” The Annals of Statistics 6(2):461-464 (1978); and Anderson et al., “Avoiding Pitfalls When Using Information-Theoretic Methods,” The Journal of Wildlife Management 66:912-918 (2002), the disclosures of which are hereby incorporated herein by reference in their entirety.
The information criteria (Table 5) are given relative to the minimum, where min AICc=90.11 and min BIC=109.5; smaller relative scores indicate better fit. BIC and AICc tend to move together, but each emphasizes different aspects of the model. BIC tends to favor models with less predictor variables (George, 2000). In the current data set, the best model by BIC is the single predictor A350 nm, and the best model by AICc is the three variable model: A350 nm, Tm″ width, and Tm′. The best two variable model is A350 nm and Tm″ width.
Converted sheet were thermoformed into retail packaging and used to assess changes in the nutritive value of specific vegetables during prolonged storage. Changes were correlated to the clarity and permeation characteristics of the plastics. All packages were stored in controlled-environment rooms equipped with 2 banks of fluorescent lights. Light intensity at the package surface was measured as Lux using a quantum light meter. Nutritive changes were assessed as decreases in Vitamin C, β-carotene, and chlorophyll during simulated retail distribution and display, up to 21 days after packaging.
Free ascorbic, dehydroascorbic (DHA), and total ascorbic (TA) acid content was determined for pre-cut Romaine lettuce using High Performance Liquid Chromatography (HPLC). Initial storage was 5 days at 3.3° C., 80% R.H. in the dark, followed by placement of half the samples at 5° C., 80% R.H under constant light, while half the samples remained at this same temperature in total darkness. The temperature/light regime was based on the observed environments for pre-cut products shipped from California to the East Coast as discussed in Zeng et al. “Growth of Escherichia coli O157: H7 and Listeria monocytogenes in Packaged Fresh-Cut Romaine Mix at Fluctuating Temperatures During Commercial Transport, Retail Storage, and Display,” Journal of Food Protection® 77(2):197-206 (2014), the disclosure f which is hereby incorporated herein by reference in its entirety. In most cases, relative humidity during shipping, distribution, and retail display is not controlled; however, it was controlled in this study to prevent excessive dehydration of tissues during storage. Packages were sampled at 0, 5, 8, 11, 14, 17, and 21 days of storage.
HPLC analysis was performed following the method reported by Chebrolu et al. “An Improved Sample Preparation Method for Quantification of Ascorbic Acid and Dehydroascorbic Acid by HPLC,” LWT-Food Science and Technology 47(2):443-449 (2012), the disclosure of which is hereby incorporated herein by reference in its entirety. A 2 g sample of tissue was extracted with an equal weight of metaphosphoric acid (MA) (3 g/100 ml DI water), using a tissue homogenizer. The sample was centrifuged at 4500 revolutions per minute for 10 minutes after which the supernatant was passed through a 0.45 μm syringe filter in preparation for HPLC analysis. Likewise, tris(2-carboxy ethyl) phosphine hydrochloride (TCEP) was used to reduce DHA to ascorbic acid to determine DHA and total ascorbic acid content of the samples. A 2 g sample was treated with an equal weight of 3 g/100 ml MA, homogenized, and centrifuged. A 300 μL aliquot was treated with 300 of 5 mmol/L TCEP, incubated for 30 min, then filtered through a 0.45 μm filter before injection on the HPLC. The sample DHA level was calculated as the difference between free AA and TA.
The HPLC used was a Shimadzu HPLC system equipped with photodiode array and C18 spherisorb column (150 mm×4.6 mm i.d. and 3 μm particle size) held at 25° C. The primary detection wavelength was 254 nm. Runs were performed isocratically using 0.01 mol/L dihydrogen ammonium phosphate (pH 2.6) as the mobile phase. Flow rate was 1 ml/min. Standards of 1 to 10 mg ascorbic acid/100 ml MA solution were prepared from a stock solution of 100 mg ascorbic acid/100 ml MA and were used to construct a calibration curve. The concentration of ascorbic, dehydroascorbic, and total ascorbic acid in the samples were expressed as mg per 100 g fresh weight lettuce. Percent recovery was determined by taking two 10.00 g samples of lettuce from a lettuce head. One sample was spiked with 0.1 g ascorbic acid. Both samples were prepared and analyzed using the procedure described above. This procedure was repeated 4 times.
Pre-cut sweet potatoes were used as a model system. They are high in β-carotene and do not contain related carotenoids, such as chlorophyll, that would require excessive processing to remove. The potatoes were stored as smooth-cut chips, julienned strips (“matchsticks”), fry-sized wedges, and ˜½″ diced pieces. Pre-cut potatoes were treated with 100 ppm active chlorine at pH 6.5 before being packaged to reduce decay during storage. Initial storage was 5 days at 3.3° C., 80% R.H in the dark, followed by placement of half the samples at 5° C., 80% R.H under constant light, and half the samples at this same temperature in total darkness. Packages were sampled at 0, 5, 8, 11, 14, 17, and 21 days of storage and unwashed potatoes were sampled at Day 0, as well.
Beta-carotene was extracted using hexane and quantified spectrophotometrically as discussed in Picha, D. H., “HPLC Determination of Sugars in Raw and Baked Sweet Potatoes.” Journal of Food Science 50(4):1189-1190 (1985), the disclosure of which is hereby incorporated herein by reference in its entirety. Peeled sweet potatoes were initially puréed using a small, table-top food processor. A 0.5 g sample of the purée was extracted for 1 min with 10 ml of HPLC-grade hexane, using a tissue homogenizer. The homogenate was filtered through Whatman #1 paper into a 50 ml volumetric and additional hexane was added to yield a final volume of 50 ml. The absorbance of 3 ml of sample was immediately read at 440 nm. A β-carotene calibration curve was prepared using 99.5% pure β-carotene. Results were expressed as mg β-carotene/100 g fresh weight. β-Carotene makes up 86 to 90% of the carotenes present in sweet potatoes, so changes at this wavelength were correlated with changes in the nutritive value of the roots.
Previous research indicated that chives, when exposed to fluorescent light, yellowed and senesced at a much faster rate than chives stored in darkness. Since the reduction in chlorophyll is so dramatic, chives were used as the model system for this work. Initial storage was 5 days at 3.3° C., 80% R.H. in the dark, followed by placement of half the samples at 5° C., 80% R.H under constant light, and half the samples at this same temperature in total darkness. Punnets were sampled at 0, 5, 8, 11, 14, 17, and 21 days of storage.
To prevent the photodegradation of chlorophyll, all glassware was covered with aluminum foil. From each sample 25 g of chives was weighed and placed in a blender. The tissue was homogenized for 1 min with 50 ml of deionized water. Ten grams of homogenized tissue was weighed into a beaker and 50 ml of 80% acetone:water (v/v) added as disclosed in Yamauchi et al., “Chlorophyll and Xanthophyll Changes in Broccoli Florets Stored Under Elevated CO2 or Ethylene-Containing Atmosphere,” HortScience 33(1):114-117 (1998) and Yamauchi et al., “Ascorbic Acid and Beta-Carotene Affect the Chlorophyll Degradation in Stored Spinach (Spinacia oleracea L.) Leaves,” Food Preservation Science 24(1):17-21 (Japan) (1998), the disclosures of which are hereby incorporated herein by reference in their entirety.
The sample was allowed to rest for 10 min. The extract was filtered through Whatman P8 filter paper using a vacuum system. Two 10 ml aliquots of 80% acetone:water were used to rinse the extraction beaker and solids remaining on the filter paper to ensure full extraction of the chlorophyll. The filtrate was further clarified by passing it through a non-reactive PTFE 0.45 μm syringe filter. Total chlorophyll was measured using a spectrophotometer set at 645 nm.
Sample Preparation:
Virgin PET resin and washed post-consumer (PC)-PET flake were blended utilizing an industrial extruder to produce 0, 25, 50, 75, and 100% PCR-PET by weight. Briefly, washed post-consumer flake was transferred via vacuum lines to a Con Air crystallizer and crystallized for 45 minutes utilizing an air temperature of 155° C. The crystallized PC-flake was blended with virgin resin in an AEC Whitlock OS series blender then transferred to a Con Air carousel drier; the blended material was dried for four hours at 140° C. with a −40° C. dew point. The dried material blend was transferred to a Reifenhäuser single screw extruder with a screw L/D ratio of 32:1 and extruded with a screw speed of 80 rpm. Film thicknesses for each sheet blend ranged between 18 mil and 20 mil (0.457 mm and 0.508 mm, respectively).
Absorption Spectroscopy:
Film samples were fitted into a custom made clamp measuring 12.7 cm×8.5 cm×2.3 cm. Ultraviolet-visible (UV-vis) spectra were collected between 200 nm and 500 nm in absorbance mode using a Tecan Safire spectrometer (Zurich, Switzerland). Specimen thicknesses were measured with a Mitutoyo IP 65 electronic digital micrometer (Kawasaki, Japan). Each specimen clamp was inserted into the sample chamber such that the specimen was aligned perpendicular to the incident irradiation. Each UV-vis absorbance value measured along the spectrum was divided by the thickness of the specimen to account for differences in path length attributed to the variations in the thickness of each specimen between specimens.
Vibrational Spectroscopy:
Attenuated total reflectance-Fourier transform infrared (ATR-FTIR) spectra were collected with a Nicolet 380 Infrared spectrometer at ambient temperature (Waltham, Mass.), fitted with a diamond crystal stage. The spectrum of each specimen was collected with 64 scans and a resolution of 2 cm-1.
Raman spectra were collected at ambient temperature on an XploRa Plus confocal Raman microscope (Horiba Scientific/JY, France) using a 100× magnification, 0.90 numerical aperture microscope objective. Each spectrum was comprised of three exposures with ten seconds per exposure and a resolution of 2 cm−1. The excitation wavelength was 532 nm with a laser power of 10 mW. Each spectrum was baseline corrected and the changes in area of each functional group characteristic band was calculated after normalization to the area of the 705 cm−1 band according to the procedure (equation 1) described in Richard-Lacroix et al., “Orientation and Structure of Single Electrospun Nanofibers of Poly(ethylene terephtalate) by Confocal Raman Spectroscopy.” Macromolecules 45:1946-53 (2012), the disclosure of which is hereby incorporated by reference in its entirety.
A(v)subtracted=A(v)% RPET−A(v)virgin RPET (1)
Fluorescence Spectroscopy:
Fluorescence intensity measurements were collected with a Tecan Safire fluorometer (Zurich, Switzerland). Scans were performed in 3D fluorescence mode with excitation and emission wavelengths ranging from 300 nm to 600 nm in 5 nm increments at a gain of 75. The z-position was held fixed at 11,020 nm.
Statistical Analysis: Investigations for linear relationships and the statistical significance of trendline slopes between data sets were performed by calculating the Pearson product moment correlation coefficient and linear regression analysis, respectively, via Minitab 17 software utilizing a 95% confidence (α=0.05) as described in Ellison, Practical Statistics for the Analytical Scientist: A Bench Guide. 2nd Ed. Cambridge: The Royal Society of Chemistry (2009), which is hereby incorporated by reference in its entirety. Statistical comparison between trendline slopes of the current data and the previous work described in Curtzwiler et al., “Effect of Recycled Poly(ethylene terephthalate) Content on Properties of Extruded Poly(ethylene terephthalate) Sheets,” J. Plast. Film Sheeting 27:65-86 (2011), which is hereby incorporated by reference in its entirety, was conducted by determining the interaction effect as described in Gelman, Data Analysis Using Regression and Multilevel/Hierarchical Models, New York, New York: Cambridge University Press (2007).
Ultraviolet Visible Spectroscopy:
Virgin and PCR-PET polymer intrinsically absorbs UV radiation at wavelengths between 280 and 320 nm, which saturated the absorption detector utilizing the sample thickness (˜20 mil), common for packaging applications as shown in
The thickness-normalized absorption at 350 nm was additionally evaluated as a potential single point measurement to predict the UVA absorption potential as defined above. Pearson's product moment correlation coefficient indicated a linear relationship between the A350 nm/mil and the UVA absorption potential for both the 2011 and 2016 data sets independently (p<0.001 and p=0.009, respectively). Regression analysis and investigation of the interaction effect of when the data were collected (2011 compared to 2016) indicated that there was no statistical difference between the trendline slopes (66.96 for 2011; 68.91 for 2016; p=0.873), and thus, values from both data sets (2011 and 2016) were combined and utilized to produce a regression equation capable of providing a single measurement investigation of the UVA absorption potential (R2=0.9289; Equation 2,
Vibrational Spectroscopy:
Understanding and determining the mechanism of increased UV absorbance would enable converters and recyclers to reproducibly tailor blend compositions and processing parameters to selectively control the ultraviolet and visible light absorption properties. Accordingly, the molecular composition of each PCR-PET blend was investigated by vibrational spectroscopic methods. Both ATR-FTIR and Raman spectra indicated an overall decrease in carbonyl concentration for specimens containing PCR as shown in
_ENREF_14 However, these spectroscopic methods do not easily differentiate carbonyl types, thus, quantification of the carbonyl decrease due to the loss of aldehyde species is confounded by the formation of quinone derivatives on the terephthalate constituents as described in Romão et al., “Poly (ethylene terephthalate) Thermo-Mechanical and Thermo-Oxidative Degradation Mechanisms,” Polymer Degradation and Stability 94: 1849-59 (2009) and MacDonald, “New Advances In Poly(ethylene terephthalate) Polymerization and Degradation,” Polymer International, 51: 923-30 (2002), which are hereby incorporated by reference in their entirety. A reduction in the asymmetric ether stretch frequency at 1095 cm−1 was also observed, which may be due to the degradation of the diethylene glycol substituents in the PET backbone (
The reduction of the carbonyl, ester, ether, and C═C functional groups as a function of PCR concentration deviated from linearity for the 50% PCR-PET blend as determined via ATR-FTIR and Raman spectroscopy (
During melt processing of the material, it is possible that the dissimilar molecular weight molar ratio between the virgin material and the PCR-PET creates heterogeneity in the material (e.g., phase separation/domains of various molecular weights or less perfect crystalline structures). Since PCR-PET has a higher crystalline fraction than virgin PET under the same extrusion/take up conditions (as measured via differential scanning calorimetry)_ENREF_21 as discussed in Curtzwiler et al., “Effect of Recycled Poly(ethylene terephthalate) Content on Properties of Extruded Poly(ethylene terephthalate) Sheets,” J Plast Film Sheeting, 27: 65-86 (2011), which is hereby incorporated by reference in its entirety, heterogeneity would result in a preferential segregation of PCR-PET forming a heterogeneous mixture of each material in the final blend, which would be most noticeable at 50 wt % PCR-PET. The resulting dispersion would result in more stability of the aromatic ring through ordered structure, and a greater crystalline fraction in the CH2 deformation vibration between 720 and 725 cm−1. Analysis of the crystalline fraction of the 50 wt % PCR-PET via ATR-FTIR revealed an increase in the concentration of the crystalline fraction over that in virgin PET resin and the other PCR-PET samples. The increase in the crystalline CH2 deformation domain supports the possibility of a heterogeneous dispersion with the 1:1 blend ratio as illustrated in
Fluorescence Spectroscopy:
Thermo-mechanical processing/reprocessing of PET induces main chain and end group degradation that results in the production of fluorescent quinone derivatives due to radical attack of the terephthalic acid aromatic ring as described in MacDonald, “New Advances in Poly(ethylene terephthalate) Polymerization and Degradation,” Polymer International, 51: 923-30 (2002), which is hereby incorporated by reference in its entirety. A strong coefficient of determination (R2=0.9878) was determined between the absorbance at 350 nm (UV-Vis spectroscopy) and the fluorescence intensity at 501 nm (ex 350 nm), suggesting that the moiety causing the increased absorbance at 350 nm (as noted in
3D fluorescence scans were utilized to identify the peak fluorescent intensity of each PCR-PET blend as shown in
Although preferred embodiments have been depicted and described in detail herein, it will be apparent to those skilled in the relevant art that various modifications, additions, substitutions, and the like can be made without departing from the spirit of the invention and these are therefore considered to be within the scope of the invention as defined in the claims which follow.
This is a divisional application of U.S. patent application Ser. No. 15/491,201, filed Apr. 19, 2017, which claims the benefit of U.S. Provisional Patent Application Ser. No. 62/324,790, filed Apr. 19, 2016, which is hereby incorporated by reference in its entirety.
| Number | Date | Country | |
|---|---|---|---|
| 62324790 | Apr 2016 | US |
| Number | Date | Country | |
|---|---|---|---|
| Parent | 15491201 | Apr 2017 | US |
| Child | 16389663 | US |
