2.1 Chemicals

Chloroform (Chromasolv, for residue analysis, ≥ 99.9%) was provided by Sigma-Aldrich (Steinheim, Germany). Diethyl ether (≥ 99.5%) was obtained from Thermo Fisher Scientific (Schwerte, Germany). n-Heptane (Chemsolute, for HPLC, ≥ 99.2%), methanol (Chemsolute, LC‒MS, > 99.9%), and methyl tert-butyl ether (MTBE, Chemsolute, for HPLC, ≥ 99.8%) were purchased from Th. Geyer (Renningen, Germany). Ultrapure water (> 18 MΩ cm) was supplied by a Millipore Synergy System (Schwalbach, Germany). Formic acid (analytical reagent grade, > 98%) was provided by Thermo Fisher Scientific. Magnesium chloride was obtained from Merck (Darmstadt, Germany), and potassium carbonate (K2CO3, anhydrous, for analysis, ≥ 99%) was purchased from Carl Roth (Karlsruhe, Germany). Primuline (dye content 50%), stearic acid (> 99.5%), 1-stearoyl-rac-glycerol (> 99%), 1,2-distearoyl-rac-glycerol (> 99%), 1,3-distearoylglycerol (> 99%), and glyceryl tristearate (> 99%) were provided by Sigma-Aldrich. HPTLC silica gel F254s MS-grade plates from Merck were used without pre-washing. A technical grade LACTEM emulsifier sample was provided by a manufacturer.

2.2 Sample and standard solutions

Sample solutions of the stored emulsifier samples were prepared at a concentration of 500 mg L−1 in MTBE. Between measurements, sample solutions were kept in the refrigerator at 4 °C. Stock solutions of stearic acid (SA), 1-monostearate (1-MSt), 1,2- and 1,3-distearate (1,2-DSt and 1,3-DSt), and tristearate (TSt) were prepared in MTBE at 1 mg mL−1. A standard mix solution was prepared at 0.2 mg lipid class mL−1 and used for comparison. For the lactylated acyglycerols, no analytical standards are available. Identification relies on a previous study [5].

2.3 Storage of a LACTEM emulsifier

The LACTEM emulsifier was stored at 60 °C in a closed Schott flask in a heating cabinet (T5042, Heraeus, Hanau, Germany). Because the LACTEM was in liquid state at this temperature, 60 °C was chosen. After distinct storage times (0 h, 1 h, 2 h, 4 h, 8 h, 24 h, 168 h (1 week), 336 h (2 weeks), 672 h (4 weeks), and 1344 h (8 weeks)) samples were collected and stored in the refrigerator prior to HPTLC analysis (Sect. 2.6). For the aerosol whipping cream manufacturing (Sect. 2.4), the same samples were used, except for the 0-h storage point, where values determined in preliminary experiments for the same emulsifier were applied.

2.4 Manufacturing of aerosol whipping cream

The aerosol whipping cream was manufactured according to [6]. In short, 0.8 g 100 g−1 emulsifier was added to pasteurized cream (lipid content of 30 g 100 g−1). This emulsifier content was found to be the saturation content in preliminary experiments. After pre-emulsification, the samples were homogenized with a two-stage homogenizer (6/1 MPa). Then, the samples were collected in flasks, cooled in ice water, and stored at 4 °C for 24 h in a plate heat exchanger. After that, rheological properties were conducted. Foaming was performed according to [11] with nitrous oxide (15 g in total).

2.5 Determination of techno-functional properties2.5.1 Particle size distribution

Particle size distribution in aerosol whipping cream was performed as described in previous publications, using static light scattering with a LS 13320 (Beckmann-Coulter, Brea, CA, USA) and a sample injection volume of 100 µL [3, 11]. For protein-stabilized fat globules, a refractive index of 1.46, as reported by [12], was applied. Characterization was based on the D90,3, the diameter that 90% of the particles do not exceed in volume-based evaluation. Each sample was analyzed three times in triplicate or quintuple.

2.5.2 Rheological properties

Rheological measurements of liquid samples were conducted on a MCR 502 and a MCR 302 (Anton-Paar GmbH, Graz, Austria) equipped with a coaxial cylindrical geometry (do = 27 mm, di = 25 mm, l = 40 mm) with a measuring gap of 1 mm. In total, 13 g of the sample was weighed into the geometry and equilibrated for 10 min at 5 °C. The shear rate profile was as follows: from 0 s−1 to 500 s−1 over 250 s, kept at 500 s−1 for 250 s, and decreased from 500 s−1 to 0 s−1 over 250 s. Apparent viscosity was calculated when the shear rate reached 500 s−1. Samples were analyzed two times in triplicate or quintuple.

2.5.3 Overrun

Overrun was calculated by differential weighing according to [11]. The density of the model aerosol whipping cream was determined to 1010 kg m−3 at 5 °C using a DMA 5000 (Anton-Paar GmbH, Graz, Austria). Samples were analyzed two times in triplicate.

2.5.4 Foam firmness

Foam Firmness was analyzed according to [3] with a universal testing machine (5944; Instron, Norwood, USA) using a crosshair probe (0.1 cm wire diameter). Foam firmness was defined as the arithmetic mean of the last 20 measurement points. Samples were analyzed two times in triplicate.

2.5.5 Drainage

Drainage was assessed according to [3]. The amount of drained serum was determined by differential weighing and normalized to the amount of foam used. Samples were analyzed two times in triplicate.

2.5.6 Residual cream

After overrun measurements and discarding of the remaining sample, residual cream was calculated according to [3] by differential weighing. Samples were analyzed two times in triplicate.

2.6 High-performance thin-layer chromatography–fluorescence detection

HPTLC coupled to fluorescence detection (FLD) were carried out as described by [6]. Briefly, 10 µL of the sample solution (resulting in 5 µg LACTEM per zone) were applied bandwise (band length 5 mm) on 20 cm × 10 cm plates with an Automatic TLC Sampler 4 (CAMAG, Muttenz, Switzerland). The distance from the lower edge was 8 mm and from the left and right edge, 10 mm. MTBE was used as the rinsing solvent with one rinsing and filling cycle. Following application, the plates were dried in a fume hood for 10 min. Subsequent twofold development was performed in an Automatic Developing Chamber (ADC2, CAMAG) equipped with a 20 cm × 10 cm twin-through chamber (CAMAG). Before each development, plate activity was set by the automatic humidity control unit of the ADC2 using a saturated magnesium chloride solution (relative humidity of 33%). The first development was carried out using a mixture of chloroform–methanol–water–formic acid (67:6:1.2:0.2, V/V) up to a migration distance of 50 mm, followed by a 10 min drying step. The two-phase solvent system was well shaken until being filled in the developing chamber briefly before the start of the development. For the second development, a mixture of n-heptane–diethyl ether–formic acid (55:45:1, V/V) up to a migration distance of 80 mm was applied, including 5 min of drying. After every working step, plate images were captured with the TLC Visualizer (CAMAG) under UV 254 nm and UV 366 nm illumination. For visualization of the LACTEM components, plates were dipped into a solution of primuline (0.05% in acetonewater, 4:1, V/V) with the TLC Chromatogram Immersion Device III (CAMAG, immersion speed: 1, immersion time: 2) and dried in a stream of cold air for 1 min. The plates were then stored in a desiccator at a constant relative humidity of 47% adjusted by a saturated K2CO3 solution for 1 h. After that, plate images were captured once more. Densitograms were obtained by scanning the plates in fluorescent mode at UV 366 nm with the TLC Scanner (CAMAG) using the mercury lamp with the optical filter being set to K400 and an analog offset of 10%. Control of the HPTLC instruments and data evaluation was performed with the software winCATS, version 1.4.6.2002 (CAMAG).

2.7 Statistical analysis

Seven characteristic LACTEM emulsifier signals were selected, and their intensities at each storage point were determined by HPTLC–FLD (n = 5). To detect significant changes in signal intensities over time and identify where significant changes occur, one-way analyses of variance (ANOVA) followed by Tukey’s honestly significant difference (HSD) test were performed at a 95% significance level. Measurement uncertainty was calculated as margin of error at the same significance level. Statistical analyses were conducted using RStudio, version 2024.09.0 (Posit Software, PBC, Boston, MA, USA), running on R, version 4.3.1 (R Foundation for Statistical Computing, Vienna, Austria). Plots were created using OriginPro, version 10.1.0.178 (OriginLab Corporation, Northampton, MA, USA).