Phenolic extraction of South African red wines

Thesis (MScAgric)--Stellenbosch University, 2019.ENGLISH ABSTRACT: Phenolic compounds are well known for their crucial role in red wine quality parameters such as colour, flavour and mouthfeel attributes. Knowledge about phenolic extraction during fermentation and possible wine practices or techniques to possibly modify phenolic content during red wine production is becoming a necessity in the wine industry to improve overall quality. In addition, the industry requires suitable, rapid, accurate and affordable monitoring tools to be able to improve and modify phenolic content during the process. In this study phenolic levels of fermenting samples were quantified through UV-Visible spectroscopy in combination with PLS calibration models. Furthermore, phenolic extraction was evaluated with batch statistical process control as a statistical monitoring tool. Eight grape batches of the commercial cultivars Cabernet Sauvignon and Shiraz were randomly selected from the Western Cape wine region. Wines were pressed at 1/3rd, 2/3rds and near the end of alcoholic fermentation and two punch down regimes were evaluated for each pressing time. Standard punch down (Cb) consisted of 3 punch downs per day during the maceration stage of skin contact, whereas increased punch down (Ta) consisted of 12 punch downs applied during different stages of fermentation. Firstly, significant variance was observed between grape batches for the four measured phenolic parameters (anthocyanins, tannins, colour density and total phenolics), with less prominent differences observed for colour density between batches. Furthermore, soft independent modelling of class analogy (SIMCA) showed wines classified according to the grape batch. Moreover, wines produced under different maceration conditions separated according to the vineyard the grapes were sourced from in the batch level model (BLM). All of the analysis indicated significant phenolic variance between grape batches. In the BLM OPLS-DA model the fermentation samples collected during maceration separated according to the different pressing times for both cultivars, regardless of punch down level. Fermenting samples obtained from pressing time 1 were associated with low phenolic content, whereas wines pressed at 2/3rds and near the end of fermentation showed trends of higher phenolic content. The BLM OPLS-DA between pressing time 2 and 3 revealed clearer trends for Cabernet Sauvignon, whereas fewer phenolic differences were observed for Shiraz samples. Fermenting samples obtained from pressing time 2 showed higher levels of anthocyanins in comparison with wines pressed near the end of fermentation were associated with higher levels of tannin, polymeric phenol and gallic acid. Shiraz fermenting samples pressed at time 3 showed only the latter. Results indicated Cabernet Sauvignon may possibly be a better suited cultivar for longer maceration, since fermentations pressed near the end of fermentation i.e. longer skin maceration, were associated with higher tannin, polymeric phenol and pigment content. In addition, results also reflected phenolic extraction as a diffusion process driven by maceration length. Furthermore, evaluating the BLM OPLS-DA a poorer separation was observed between standard and increased punch down frequency. Increased punch down frequency showed trends of higher phenolic content, whereas fermenting samples produced with three punch downs a day was strongly associated with high phenolic acid content only. Phenolic differences were observed between Cabernet Sauvignon and Shiraz indicating results may be batch or variety dependent. Additionally, taking a closer look at both punch down levels for each pressing time, similar results were observed in the OPLS-DA models. Overall, BSPC allowed for the monitoring of phenolic extraction and identification of possible deviations during maceration. However, this study data was only evaluated after process completion. This methodology could be potentially used to monitor phenolic extraction in real time for future red wine fermentations. Additionally, the approach of PLS calibrations proved to be a suitable, rapid, accurate and cost-effective method to measure phenolic levels of fermenting samples.AFRIKAANSE OPSOMMING: Fenoliese komponente is veral bekend vir hulle belangrike rol in rooiwyn kwaliteit parameters soos kleur, geur en mondgevoelskenmerke. Kennis oor fenoliese ekstrasie gedurende fermentasie en moontlike wynmaak praktyke of tegnieke wat moontlik fenoliese ekstrasie kan modifiseer gedurende rooiwyn produksie word al hoe meer ‘n noodsaaklikheid in die wyn industrie om algehele rooiwyn kwaliteit te verbeter. Boonop vereis industrie geskikte, vinnige, akkurate en bekostigbare moniteringsinstrumente om kwaliteit te verbeter, sowel as fenoliese inhoud gedurende die proses te modifiseer. Gedurende die studie is fermentasie monsters se fenoliese vlakke gekwantifiseer deur UV-Sigbare spektroskopie in kombinasie met PLS kalibrasie modelle. Fenoliese ekstrasie was geevalueer met statistiese prosesbeheer as ‘n statistiese moniteringsinstrument. Agt groepe druiwe van die kommersiële kultivars Cabernet Sauvignon en Shiraz is geselekteer vanuit die Wes-Kaap wynstreek. Gedurende alkoholiese fermentasie was wyne gepers by 1/3rde, 2/3rdes en naby die einde van alkoholiese fermentasie, sowel as twee deurdruk tegnieke/frekwensies was geëvalueer vir elke periode waarby die druiwe gepers was. Standaard deurdrukke (Cb) het bestaan uit drie deurdrukke per dag gedurende die maserasie tydperk waartydens die doppe in kontak was met die sap, terwyl verhoogde deurdrukke (Ta) bestaan het uit 12 deurdrukke per dag gedurende die verskillende fases van fermentasie. Eerstens was beduidende verskille waargeneem tussen die verskillende druif groepe vir die vier fenoliese parameters gemeet naamlik antosianiene, tanniene, kleurdigtheid en totale fenole, met minder prominente verskille tussen druif groepe vir kleurdigtheid. Die analise sagte onafhanklike moddelering van klas analogie (SIMCA) het aangedui dat wyne geklassifiseer was volgens die druif groepe waarvan die fermenterende monsters oorspronklik van afkomstig was. Boonop het die wyne wat geproduseer was tydens verskillende maserasie kondisies in die bondelvlakmodel (BLM) geskei volgens die wingerd waarvan die druiwe afkomstig van was. Alle analises het daarop gedui dat daar beduidende verskille tussen die druiwe was ten opsigte van fenoliese komposisie. In die BLM OPLS-DA model het die fermentasie monsters geskei volgens die verskillende maserasie tyetoe die wyne gepers was, onafhanklik van die deurdruk frekewensie. Fermentasie monsters wat tydens perstyd 1 gedruk was, was geassosieerd met lae vlakke van fenoliese inhoud, terwyl wyne wat gepers was by 2/3rdes en naby aan die einde van alkoholiese fermentasie het tendense getoon van hoë vlakke van fenoliese inhoud. Die BLM OPLS-DA het meer tendense getoon tussen wyne van perstyd 2 en 3 vir Cabernet Sauvignon, terwyl minder tendense waargeneem is vir Shiraz monsters. Fermentasie monsters van perstyd 2 het tendense getoon van hoër vlakke antosianiene in vergelyking met wyne gepers naby aan die einde van alkoholiese fermentasie wat geassosieerd met hoër vlakke van tannien, polimeriese fenol en gallinsuur. Shiraz fermentasie monsters gepers by tyd 3 het slegs tendense getoon van die laasgenoemde. Resultate dui daarop dat Cabernet Sauvignon moontlik ‘n meer geskikte kultivar is vir verlengende maserasie kondisies, aangesien fermentasie monsters gepers naby die einde van alkoholiese fermentasie, m.a.w. verlengde dop maserasie, geassosieerd was met hoër tannien, polimeriese fenol en pigment inhoud. Boonop toon resultate dat fenoliese ekstraksie ‘n diffusie proses is gedryf deur die maserasie-lengte. Na evaluering van die BLM OPLS-DA, was ‘n swak skeiding tussen standaard en verhoogde deurdruk frekwensie geobserveer. Verhoogde deurdruk frekwensie het tendense getoon van hoër vlakke van fenoliese inhoud, terwyl fermentasie monsters geproduseer met 3 deurdrukke per dag het slegs tendense getoon van hoër fenoliese suur inhoud. Fenoliese verskille was waargeneem tussen Cabernet Sauvignon en Shiraz wat aandui dat resultate dui dat resultate kultivar of groep afhanklik kan wees. Daarbenewens, evaluering van beide deurdruk frekwensie gedurende die verskillende tye van pers was soortgelyke resultate in die OPLS-DA resultate gevind. Oor die algemeen het BSPC die monitering van fenoliese ekstraksie en identifikasie van moontlike afwykings gedurende maserasie toegelaat. Die studie data was egter slegs geëvalueer na proses voltooïng. Die metode kan potensieël gebruik word om fenoliese ekstrasie intyds van toekomende rooiwyn fermentasies te monitor. Daarbenewens, die gebruik van PLS kalibrasies het bewys dat dit ‘n geskikte, vinnige, akkurate en koste-effektiewe metode is om fenoliese vlakke van fermentasie monsters te kwantifiseer.Master