Chapter 4: Bloom identification & mechanisms – untempered chocolate
Effects of fat polymorphic transformation and non-fat particle size distribution on the surface changes of untempered model chocolate, based on solid cocoa mass 1
The results reported in this chapter can facilitate the understanding of fat bloom formation in untempered chocolate with respect to the changes in microstructure and surface appearances. It also contributes to show the details of IV-to-V polymorphic transformation in the fat phase as time went by.
Plain (dark) chocolate is primarily composed of cocoa butter and non-fat particles including sugar crystals and cocoa powder . Manufacturing processes of commercial chocolate generally involve mixing, refining, tempering, and conching . Since cocoa butter has six polymorphic forms, I to VI in order of increasing polymorphic stability , proper tempering is critical for liquid chocolate to crystallise into the right polymorph (form V) giving the products a long shelf-life of more than one year . However, chocolate can experience a reduced shelf-life due to “fat bloom” caused by reasons such as long-term storage, a failure of tempering, and exposure to high temperatures .
The most frequently reported fat bloom, sometimes called storage bloom, refers to the whitish appearance that forms on properly tempered chocolate during long-term storage due to the separation and deposition of small fat crystals on the chocolate surface [90, 92, 111, 123]. However, bloom is a general term that describes a loss of initial surface gloss during storage and has varying definitions under different circumstances . One important type of fat bloom forms on the surface of untempered/under-tempered chocolate which often contains a large proportion of form IV crystals . These unstable form IV crystals are present due to a lack of form V nuclei during crystallisation or recrystallisation of molten chocolate which may occur during improper handling and storage . Form IV crystals are soft, and spontaneously transform into form V crystals in a period of days at room temperature. This causes pronounced bloom leaving mixed patches of light and dark areas on the chocolate surface .
Studies have characterised compositional differences on the surface of untempered/under-tempered chocolate after blooming. Lonchampt and Hartel  reported that the lighter coloured areas appeared whitish and were essentially sugar and cocoa particles based on the melting enthalpies of sugar and fat in samples collected from the light areas. This is supported by Kinta and Hatta  who used elemental analysis to determine that light areas had less fat than dark areas, but the light areas were not totally devoid of fat. However, based on microstructural observations of bloomed under-tempered chocolate, Afoakwa et al.  argued that the light areas were mixtures of fat and sugar crystals while the dark areas were primarily cocoa particles. They assigned the resultant whitish appearance to recrystallisation of the migrated fat on light areas.
The driving force of bloom development on untempered chocolate has been proposed to be crystal formation around the form V crystals leading to the redistribution of fat content [44, 92], and Afoakwa et al.  added that liquid fat movement to the chocolate surface was initiated by capillary action and continued by diffusion due to a concentration gradient of fat.
Due to contradictions in the literature related to the identification and mechanism of bloom further studies are needed to clarify the surface structure of untempered chocolate. In particular, to understand the different observations found by Kinta and Hatta  and Lonchampt and Hartel  with the former reporting dark spots rich in fat surrounded by light areas on bloomed chocolate, and the latter observing white bloom spots in the fat-rich dark areas. Moreover, there is little information available on the relationships between polymorphic transformation and the changes in surface appearance during the blooming process. Therefore, the objective of this study was firstly to relate the polymorphic transformation of cocoa butter to the associated surface changes during bloom formation using a simple cocoa mass model. Then the microscopic surface morphology was characterised after blooming. Finally, the particle size effect on bloom formation was also studied.
Materials and Methods
Four model chocolates were prepared based on cocoa mass (CM) with differing particle size distributions. The CM chocolate was prepared by grinding roasted cocoa nibs (Chocolate Brown, Warkworth, NZ) in a chocolate melanger (Spectra 11 stone melanger, Coimbatore, India). For each of the four batches, 600g of cocoa nibs were pre-ground at 700rpm for 40min. The four different particle size distributions were obtained by further grinding the liquor at 1200rpm for either 3, 5, 7, or 10 hours.
Fat removal and particle size measurement
Direct hexane extraction  was used for fat removal in CM chocolate. Approximately 5g of CM chocolate was dissolved in 20ml of n-hexane (>98.5%). Sample was left for 24 hours to allow the cocoa particles to settle at the bottom of the beaker. The upper hexane layer in which the cocoa butter was dissolved was carefully removed using a pipette. This procedure was repeated a total of three times using fresh n-hexane for each extraction. Then the cocoa particle layer was dried in a fume hood for 2 hours until all of the n-hexane had evaporated. To separate the particles, the cocoa particle layer was lightly crushed using a pestle and mortar.
The particle size distribution of the cocoa particles obtained was measured using a laser diffractometer (Mastersizer 2000, Malvern Instruments Ltd., Malvern, U.K.). Approximately 0.1g of powdered sample was dispersed in water and ultrasound treated for 2min prior to measurements. The refractive index of the particles was set to 1.59 (as determined experimentally in Appendix B) and a pump speed of 200/min was used. Each type of cocoa particle was measured in triplicate. Particle size parameters including D90 (90%<the size), D50 (50%<the size), D32 (the Sauter mean diameter) and D43 (the DeBroukere mean diameter) were then obtained, among which the D90 values were used to represent the four batches of cocoa particles since it is commonly used in the chocolate industry to indicate the proportion of the large particles [8, 25].
Each of the CM chocolates was melted in a 60oC water bath (WB-11, WiseBath, South Korea). The molten chocolate was continuously stirred for 1 hour at 200rpm using a mechanical stirrer (RW 20 digital, IKA-works Inc. NC, USA) to ensure that all crystal history was erased. Subsequently, the liquor was poured into cylindrical moulds (diameter: 4.5cm and depth: 1.2cm) and sealed in a container before cooling in a refrigerator (5±0.5oC) for 1 hour. The containers were then removed from the refrigerator and stabilised at ambient temperature (20±0.5oC) for 2 hours. Afterwards, sample discs were demoulded and stored at ambient temperature with relative humidity never exceeding 50%. Each type of CM chocolates had five sample discs, which were run in parallel for different measurements over 25 days of storage.
Surface image acquisition and processing
Every 24 hours, the same areas on sample surfaces were imaged using a light microscope (Nikon AZ100 multizoom microscopy). At least four locations on the surface of a sample disc were examined; three sample discs of each CM chocolate were used. The low magnification images obtained were adjusted in Adobe Photoshop CS6 (Adobe Systems Inc., San Jose, CA, USA) where the images were converted to grayscale, and optimum brightness and contrast were selected. ImageJ software  was then used to calculate the area ratio of surface bloom on chocolates stored for 25 days, i.e. white/dark area ratio.
X-ray diffraction (XRD) is commonly used to examine the polymorphism of cocoa butter, but one issue that complicates the use of XRD of chocolate is that diffraction from sugar crystals obscures cocoa butter diffraction peaks . However, the CM chocolate was formulated without sugar. Therefore, in this study, the polymorphic transformation of cocoa butter was examined directly using XRD (D2 Phaser, Bruker, Germany). A small piece with a flat surface was cut from each type of sample discs using a scalpel. It was scanned using Cu-Kα X-rays (goniometer step size = 0.02°, dwell time = 0.1s) and the diffraction pattern was recorded.
A machine vision system  was used to measure the changes in surface Whiteness Index (WI) during bloom, and the “two images” method , which eliminates the influence of background colour, was employed. The images were analysed using LensEye software (Gainesville, FL, USA) and the WI values were automatically displayed. The WI values were based on L* a* b* (colour space) values by the software using Equation 3 (page 57) and normalised WI was calculated by Equation 4 Where WIt = whiteness index at time t and WI0 = initial whiteness index.
At the end of bloom formation, the surface microstructure of samples cut from the discs was imaged in its native state in low vacuum conditions using Environmental Scanning Electron Microscopy (ESEM) (Quanta 200 FEG SEM, FEI Company, Eindhoven, The Netherlands). The sample was cooled to a temperature of -5°C on a peltier-cooled stage to reduce thermal damage from the electron beam , and imaged in a nitrous oxide atmosphere at a pressure of 77.3Pa and an accelerating voltage of 20kV. Samples were also viewed using cryogenic SEM (cryo-SEM; XL30S FEG SEM, The Netherlands) equipped with an Alto 2500 cryo-chamber (Gatan UK, Abingdon, UK). A piece of sample was plunge frozen in liquid nitrogen slush and sputter coated with gold prior to imaging.
A thin piece of sample was cut from the surface of bloomed chocolate and mounted on a fluorescence microscope (Nikon Eclipse 80i) equipped with a DS-Fi1 (Nikon) digital camera. Images were then captured using DIC optics in dark field, using a FITC filter (excitation: 465-495nm; dichroic mirror: 505nm; barrier filter: 515-555nm). The images were processed using Adobe Photoshop software (CS6, Adobe Systems Inc., San Jose, CA) to merge areas in-focus as there were areas out of focus due to uneven sample surface. Optimum brightness and contrast were also selected.
The values of size parameters and the particle size effect on the area ratio of surface bloom were both evaluated by one-way ANOVA with a Tukey post hoc test, at a significance level of 95% (IBM SPSS Statistics for Windows, version 21.0, IBM Corp., Armonk, NY, USA).
Results and Discussion
Bloom formation during storage
Evolution of surface appearance
The four types of CM chocolates had D90 values of approximately 18, 25, 34 and 56µm, respectively. They showed similar surface changes when observed either visually or by the light microscope; hence, only the results of 18µm CM chocolate are shown in Figure 4-1 and Figure 4-2.
When the chocolates were demoulded, the surface was brown and with a relatively homogeneous appearance (Figure 4-1, day 1). However, it was observed that within several hours after demoulding small spots developed, initially appearing light brown; afterwards, the number and area of the spots increased during the first 2 days of storage. As seen in Figure 4-1, from day 3, the bloomed area which was the white periphery surrounding the spots started to appear. After approximately 6 days of storage, the peripheral area was more pronounced and had joined together making up the majority of the surface colour, while the original brown colour had almost disappeared. From approximately 10 days of storage the surface stopped changing.
The process of bloom formation was also investigated using light microscopy by following the development of spot (S) and periphery (P) areas (Figure 4-2). During the first 3 days of storage, the light brown spots appeared as small dark nuclei surrounded by a lighter “halo” and grew in size (S1 to S3). From days 3 to 6, these lighter coloured spots darkened (S3 to S6), and the white and sandy peripheral areas (P3 to P6) rapidly expanded into the surrounding area that appeared to be brown. Thereafter, changes in spot appearance were minimal but the peripheral areas surrounding the spots became whiter. From approximately 10 days of storage, the chocolate surface had a “honeycombed” appearance composed of irregular sub-units of dark spots. These findings are in agreement with the changes shown in Figure 4-1, which suggests that this type of bloom could be accurately identified using the naked eye. In addition, a similar “honeycombed” appearance after bloom formation was found by Kinta and Hatta  using untempered chocolate and Afoakwa et al.  using under-tempered chocolate.
The polymorphic transformation during the bloom process is shown in Figure 4-3. At the start of storage, the samples had a diffraction pattern indicating the presence of cocoa butter polymorphic form IV with two strong peaks at around 4.34Å and 4.15Å. These are in agreement with characteristic peak positions reported in literature [37, 43, 45, 86]. During the first 6 days of storage, the diffraction peaks of polymorphic form IV gradually disappeared, and the characteristic peaks of form V appeared: a very strong peak at 4.59Å, a distinctive weaker peak at around 3.99Å, and three other peaks at approximately 3.89Å, 3.77Å, and 3.67Å. These are also similar to the peak positions reported for polymorphic form V in the above-mentioned literature. From day 6, the peak intensity of form V became stronger and the characteristic peaks sharpened as an indication of post-crystallisation processes such as “Ostwald ripening” (growth of large crystals at the expense of smaller crystals) .
This study revealed that the evolution of the chocolate surface and the polymorphic transformation of cocoa butter are related. Within 6 days of demoulding, the most significant changes on the chocolate surface occurred with the spots and sandy periphery rapidly developing (Figure 4-2). During the same period, most of the form IV cocoa butter transformed to form V (Figure 4-3). Therefore, although it has been generally accepted that unstable crystal formation causes fat bloom in untempered chocolate , this study suggests and provides evidence that the polymorphic transformation of form IV to V is responsible for this type of bloom.
Microstructure of bloomed chocolate
Characterisation of the microstructure
Kinta and Hatta  observed that no fat crystals appeared on the surface of untempered chocolate after bloom formation. However, in the current study separated fat crystals were found in the dark spots after 25 days of storage in Figure 4-4 (Area b: b1 and b2). The superficial structure of fat crystals showed similarities to the fat bloom found on well-tempered chocolate, i.e. needle- or flake-like crystals . However, the X-ray diffraction pattern of untempered chocolate at day 25 (Figure 4-3) revealed that the cocoa butter was still mainly in form V rather than the most stable form VI often found in bloomed well-tempered chocolate . This suggests that the formation of fat crystals on the dark spots is not due to V-to-VI polymorphic transformation of cocoa butter. It is also seen in Figure 4-4 (b1 and b2) that the fat crystals were densely packed on the surface of the chocolate. James and Smith observed fewer fat crystals that were less densely packed on the surface of untempered chocolate. However, James and Smith  used both a different chocolate and sample preparation methods.
When viewed using fluorescence microscopy, periphery areas that appear sandy with the naked eye appeared brighter relative to areas that appear as dark spots with the naked eye (Figure 4-4). The only component of the chocolate samples that can autofluoresce are the cocoa particles . Therefore, these images indicate that the periphery areas (Area a) are mainly composed of cocoa particles. This conclusion is supported by the cryo-SEM images (Figure 4-4) which show individual cocoa particles (a2) and voids between particles (a3) on the sandy periphery. The spots were darker because of the coverage of cocoa butter on the cocoa particles. As a result, the dark areas were mainly fat-based while the light areas had a lack of fat, which is consistent with the literature [68, 75]. However, it is not in line with Afoakwa et al.  who stated that the dark spots on under-tempered dark chocolate was mainly composed of cocoa particles (solids), while the periphery is mainly fat and sugar. Those statements were mainly based on observations using low magnification stereoscopic binocular microscope, which may not have provided sufficient resolution, nor can that technique identify the composition of components as fluorescence microscopy did in this current study.
Formation of the microstructure
Based on the above observations, it is clear that during bloom formation fat was redistributed from the peripheral areas towards the dark spots. Moreover, since the honeycomb-like structure was also found inside the samples as shown in Figure 4-5, it indicates that the continuous structural changes occurred uniformly throughout the entire sample, not only on the sample surface.
Table of Contents
Chapter 1 : Introduction
Chapter 2 : Literature review
2.1 : Chocolate systems
2.2 : Chocolate manufacturing
2.3 : Bloom on chocolate
2.4 : Factors influencing bloom formation
2.5 : Bloom-related properties
2.6 : Conclusions from the literature
Chapter 3 : Materials and Methods
3.1 : Development of chocolate samples
3.2 : Established methods
Chapter 4 : Bloom identification & mechanisms – untempered chocolate
4.1 : Introduction
4.2 : Materials and Methods
4.3 : Results and Discussion
4.4 : Conclusions
Chapter 5 : Bloom identification & mechanisms – well-tempered chocolate
5.1 : Introduction
5.2 : Materials and Methods
5.3 : Results and Discussion
5.4 : Conclusions
Chapter 6 : Microstructural factors affecting fat bloom formation – PSD
6.1 : Introduction
6.2 : Materials and Methods
6.3 : Results and Discussion
6.4 : Conclusions
Chapter 7 : Bloom-related chocolate properties
7.1 : Introduction
7.2 : Materials and Methods
7.3 : Results and Discussion
7.4 : Conclusions
Chapter 8 : General discussion
8.1 : Applicability of Testing Methods
8.2 : Fat bloom classification
8.3 : The relationships between microstructure, fat bloom and physical property
Chapter 9 : Concluding remark
GET THE COMPLETE PROJECT
Fat bloom in chocolate Bloom identification and its corresponding effects on chocolate properties