Background

Researchers from previous studies have reported an extensive list of published descriptors for different volatile chemicals commonly found in beef products. Since then, other studies have been successful in correlating the volatile aroma chemical compounds with trained descriptive sensory panel scores as well as consumer liking panels.

Data from a previous Checkoff‐funded project, as a proof of concept, shows a measured increase in some Maillard products (namely pyrazines) and a corresponding decrease in lipid‐degradation products like aldehydes, ketones, as well as sulfur products. The current study has demonstrated basic differences in volatile production given 3 different subprimal steaks, but the range in grilling temperature does not allow the fitment of sufficient data points to determine linear and/or quadratic impacts. Additionally, the current research as well as that in an exhaustive literature search has failed to provide any information in regards to the production of non‐volatile meat products.

Literature searches have also failed to produce a thorough description of flavor formation in intact beef except for those that utilize beef flavorings or theoretical models. The intent of this study is to produce a model system using strip loin steaks at 5 different grill temperatures to confirm that it fits existing models.

Therefore, the perceived outcome of this research is to develop a model system that will describe the production of volatile and non‐volatile compounds produced at 5 different grill temperatures. When coupled with a descriptive trained sensory panel, this will produce both a practical outcome of furthering understanding regarding the correlation of flavor chemistry with sensory descriptors as well as a theoretical outcome of advancing knowledge of flavor chemistry.

The objectives of this study were to 1) determine the impact of 5 different grill surface temperatures on the trained sensory panel scores, volatile aroma compounds, and non‐volatile compounds, 2) develop predictors of trained sensory panel scores using cooking parameters and volatile aroma compounds, and 3) develop a novel and innovative method for measuring non‐volatile flavor factors using metabolomics.

Methodology

Thirty Select strip loins were selected from carcasses at a commercial major packing plant in Texas. After selection and vacuum‐packaging, the loins were transported to the Rosenthal Meat Science and Technology Center where the strip loins were aged for 14 days. After aging, the loins were divided into 10 portions (5 grill temperatures for each of trained sensory panel and Warner‐Bratzler shear force). Steaks were labeled, vacuum‐packaged individually, and placed into frozen storage until analyses were performed. Individual steaks were selected and thawed in refrigerated (4°C) storage for 12 to 24 hours. Grills were pre‐heated to 1 of 5 different temperatures: 149°C (300°F), 177°C (350° F), 204°C (400°F), 232°C (450°F), or 260°C (500°F). Steaks for all cooked analyses were placed on a 2.5‐cm‐thick flat top Star Max Electric Griddle.

An expert trained beef flavor descriptive attribute panel with over 200 h of training and 10 y of experience consisted of six panelists trained on ten basic flavors and five texture attributes from the beef lexicon for 6 d prior to testing. Panelists were trained to scale each attribute on a 16‐point intensity scale (0 = none and 15 = extremely intense).

The extra, still warm, cubes (1.3 cm x 1.3 cm x steak thickness) from the trained panel were frozen in liquid nitrogen immediately after cutting and stored in aluminum foil at ‐80°C. For analysis, the samples were weighed and placed in a 473 mL glass jar with a Teflon lid to be placed in a water bath held at 60°C. After thawing and equilibrating for 60 min, a solid‐phase micro‐extraction (SPME) portable field sampler was injected into the injection port of a gas chromatograph. Frozen samples were extracted and each sample was run in duplicate using an Liquid chromatography– mass spectrometry/quadropole time of flight (LC/MS‐QTOF).

Data were analyzed to determine whether grill surface temperature acted in a linear or quadratic fashion on the sensory and volatile aroma compounds. Data were also subjected to analyses to determine relevant associations with each other and the ability to predict sensory traits.

Findings

Beef identity, brown, roasted, burnt, umami, and bitter flavor descriptors increased linearly as the temperature of the grill increased. Juiciness and Warner‐Bratzler shear force were not affected by grill temperature. On the other hand, both myofibrillar tenderness and connective tissue amount scores decreased linearly as grill surface temperature increased. This indicates that beef loin steaks became tougher as grill temperature increases.

Because the pyrazine classification of compounds is a product of the Maillard reaction, the pyrazines, which generally have roasted, coffee, and nutty aromas, increased as grill temperature increased. One exception was 2,6‐ dimethyl pyrazine (meaty or potato aroma) which increased quadratically with grill temperature. In fact 2,3,5‐trimethyl‐6‐ethyl pyrazine, 2,3‐diethyl‐5‐methyl pyrazine, 2,5‐dimethyl pyrazine and 3,5,‐diethyl‐2‐methyl‐pyrazine were only present when the grill temperature reached 500°F. Like the pyrazines, the ketones (fruity and fatty aromas) also increased as grill temperature increased. Finally, a pyridine was present at the highest grill temperature and styrene was only present in the lowest grill temperature, and 4‐ethyl‐2,5,6‐trimethylpyrimidine increased linearly with grill temperature.

In the upper left quadrant, the lipid degradation products like hexanal, pentanal, and 2‐butanone and trained sensory panel traits of liver‐like, sour, metallic, and bloody‐serumy are associated with the lower grill surface temperatures. The upper right quadrant, with the 500°F grill surface temperature, contains bitter and burnt sensory panel traits along with pyrazines which are a product of the Maillard reaction. Interestingly, the trained sensory panel traits of umami, beef identity, roasted, and brown seem to be pulled down into the lower right quadrant by the 450°F grill surface temperature treatment.

The prediction of beef identity included grill temperature at the flip, internal temperature and cooking time on side two as some of the most important factors that predict beef identity. Additionally, numerous pyrazine and aldehydes are also responsible for predicting beef identity and collectively account for 59.3% of the variation in beef ID. Twenty‐three cooking parameters and volatile compounds accounted for 79.2% of the variation in the brown sensory trait. Cooking parameters and volatile compounds accounted for 68.7% of the variation in roasted sensory.

Regarding non‐volatile compounds, betaine is a naturally occurring compound that has been of interest for its role in osmoregulation as a methylating agent and increases in intensity as grill temperature increased to 450°F. The positive ion 4,8,12‐trimethyl‐tridecanoic acid showed a dramatic increase in intensity when the grill temperature reached 500°F. Trimethyltridecanoic acid is the presumed breakdown product of the beta‐oxidation of pristanic acid. There was a tendency for free stearic acid to increase as the grill temperature increases, with a great deal of unknown variation in intensity of this free fatty acid which would be expected as the sample extraction process for high performance liquid chromatography (HPLC) includes a washing step which removes lipids. The negative ion (S)‐N‐(4,5‐ Dihydro‐1‐methyl‐4‐oxo‐1H‐imidazol‐2‐yl)alanine) was positively affected by the increase in grill temperature, while 2‐ Dimethylamino‐5,6‐dimethylpyrimidin‐4‐ol also showed an increase with grill temperature and an especially large increase in intensity at the 500°F grill temperature treatment. There was an increase in intensity of 4‐Hydroxy‐2',3,5,5'‐ tetrachlorobiphenyl. This increase is shown with the exception of the 450°F grill temperature treatment which decreases from the 400°F grill temperature treatment. Finally, the negative ion recovery of the compound with the formula of C9H15N3O also shows a steady, linear increase with an increase in grill surface temperature.

Implications

Increasing the grill surface temperature when cooking USDA Select beef strip loin steaks increases many of the positive trained sensory panel attributes and volatile aroma compounds important to the acceptance of beef, but decreases beef tenderness. Additionally, it appears that the cooking parameters along with volatile aroma compounds account for as much as 65% of the variation in positive beef flavor traits. The increase in grill temperature positively influences beef flavor and the metabolites that are precursors to flavor compounds in lean beef. It appears that cooking at a temperature of 400°F may maximize umami, beef identity, and roasted flavors.