*3.5. Microstructure*

The microstructures of the stirred ADG samples are shown in Figures 2 and 3 (SEM) and Figures 4 and 5 (CLSM).

(**a**) (**b**)

(**c**) (**d**) 

**Figure 4.** *Cont*.

(**e**) 

**Figure 4.** CLSM (confocal laser scanning microscopy) images of plain ADG0% (**a**), ADG with 15% black tea infusion (**b**), ADG with 30% black tea infusion (**c**), ADG with 45% black tea infusion (**d**), ADG with 60% black tea infusion (**e**) (day 1). Protein stained by Fast Green FCF appears as green and non-fluorescent areas (dark areas) correspond to the serum pores.

(**a**) (**b**)

 **5.**

*Cont*.

**Figure**

(**e**) 

**Figure 5.** CLSM images of plain ADG0% (**a**), ADG with 15% black tea infusion (**b**), ADG with 30% black tea infusion (**c**), ADG with 45% black tea infusion (**d**), ADG with 60% black tea infusion (**e**) (day 28). Protein stained by Fast Green FCF appears as green and non-fluorescent areas (dark areas) correspond to the serum pores.

SEM micrographs showed the 3D (including z-depth) organization of protein gels (stirred ADG samples). In Figure 2, comparing the morphology and the segregated structure between ADG0% and ADG15%, the z-depth structure is more densely packed in the ADG15% (Figure 2a vs. Figure 2b). Also, the later had slightly finer protein arrangement, as shown in Figure 2, ADG0% had relatively more large cavities (white arrows in Figure 2a) than those in ADG15% (Figure 2b). The structure of ADG30% (Figure 2c) was relatively thicker and denser among other gel samples on day 1. Figure 2c showed that the sample had fine protein arrangemen<sup>t</sup> resulting very small pores in the gel structure. Such microstructure explains the better phase stability (WHC, Table 1) in the gel sample containing 30% BTI. Generally, ADG15%, ADG30% and ADG45% three samples had similar spongy-like interior with few air cells and highly branched-structure. Fiszman, et al. [60] found that the smooth bridge with double network structures of dairy gel seemed to be located at the inside of network of casein micelles that could maintain the aqueous phase more effectively and reduce EOS. Although the authors did not study the impact of tea infusion on dairy protein gelation, the work clearly demonstrated that smaller pore sizes within the 3D gel network resulted in lower syneresis rate. Such observation about the relation between 3D structure of gel and its phase stability is in a good agreemen<sup>t</sup> with our observations. For instance, including even a relatively small volume of BTI may result in relatively smaller pore sizes for acidified milk gel (ADG15%, Figure 2b, white arrow) than the control gel (Figure 2a, white arrow); consequently, such structure may resulted in relatively lower EOS for ADG15% on day 1 in comparison with the control gel (Table 1). Although the impact of tea addition to yogur<sup>t</sup> (or ADG) on the gel texture, antioxidant activity, lactic acid bacteria and quality related characteristics have been studied elsewhere [15,22,27,61,62], few research has been done to investigate the impact of tea infusion on microstructure of acidified milk protein gel systems. However, other phenolic compounds rich materials have been incorporated in yogur<sup>t</sup> gels and the microstructure of the gels was studied. For instance, a recent research showed that 3% apple pomace in stirred yogur<sup>t</sup> resulted in more compacted protein network and larger cavities in the gel in comparison with negative control gel [63]; moreover, Pan and co-workers found that inclusion of 5% pomegranate juice powder in set yogur<sup>t</sup> resulted in denser protein gel packing and inhomogeneous size distribution pores [64,65]. These observations in literature are in good agreemen<sup>t</sup> with results found in this research. Some aggregated and/or lumpy protein networks (white circles in Figure 3d,e) appeared in gels with high incorporation rate of BIT (45% & 60%) after 28 days of cold storage; these structural features cannot be observed in the same samples

on day 1 (white circles in Figure 2d,e). These changes in structural organization of ADGs explain the instability of ADG45% and ADG60% (EOS and WHC results in Table 1).

CLSM was used for characterizing laminar microstructure of ADG samples. On day 1, ADG30% (Figure 4c) and ADG45% (Figure 4d) had denser protein gel blocks as pointed by white arrows. By comparing between Figure 4d,e and Figure 5d,e, more and bigger pores appeared in samples containing large volume of BTI (45% and 60%) after 28 days period of storage. The results suggested that significant transformations of gel structure took place during storage in which incorporation rate of BTI is high (45% and 60%). The structure characteristics of ADG30% remained nearly the same on day 1 and day 28. Such result is consistent in both SEM (Figures 2c and 3c) and CLSM (Figures 4c and 5c) images. The poorest gel network was found in ADG60% after 28 days storage, such structure resulted in worst cohesiveness among all gel samples (Table 2, *p* < 0.05).

In general, the structural changing trend was consistent between SEM images and CLSM images. The interactions between polyphenols and dairy proteins may be responsible to the microstructure changes of acid induced milk protein gel. The impact of tea extract components on the acidified gelation process is rather complicated as they a ffect both micro- and macro- structures at the same time.
