*2.3. Statistical Analysis*

Analysis of variance was conducted by one-way ANOVA followed by the Tukey test using Minitab® 18 (Minitab Inc., State College, PA, USA). Differences at *p* < 0.05 were considered to be statistically significant. For the agglomerative hierarchical cluster (AHC) analysis, the 12 essential oil compositions were treated as operational taxonomic units (OTUs), and the concentrations (percentages) of 18 of the most abundant essential oil components (curzerenone, curdione, germacrone, *ar*-turmerone, 1,8-cineole, α-turmerone, unidentified (RI = 1778), β-turmerone (=curlone), β-sesquiphellandrene, α-zingiberene, *iso*-curcumenol, curcumenone, *trans*-β-elemene, *ar*-curcumene, β-pinene, curcumenol, camphor, and curzerene) were used to determine the chemical associations between the *Curcuma* rhizome essential oil samples using XLSTAT Premium, version 2018.1.1.62926 (Addinsoft, Paris, France). Similarity was determined using the Pearson correlation, and clustering was defined using the unweighted pair-group method with arithmetic mean (UPGMA).

### **3. Results and Discussion**

#### *3.1. Chemical Composition of Curcuma Rhizome Essential Oils*

The fresh rhizome samples were hydrodistilled to give colorless or pale-yellow essential oils in yields ranging from 0.41% to 1.13% (Table 1). The chemical composition of the *Curcuma* rhizome essential oils are compiled in Table 2. Gas chromatograms of each Curcuma variety are shown in Supplementary Figure S1.

The essential oils from the green-colored mother and daughter rhizomes of *C. aromatica* (CA22) were dominated by curzerenone (18.6% and 14.7%, respectively), germacrone (14.7% and 10.7%, respectively), 1,8-cineole (11.7% and 6.6%, respectively), and an unidentified sesquiterpenoid (RI 1778, Supplementary Figure S2) (9.0 and 8.7%, respectively). Similarly, the white-colored mother and daughter rhizomes (*C. aromatica*, CA46) were dominated by the same components, curzerenone (14.9% and 14.8%), germacrone (14.5% and 12.5%), 1,8-cineole (10.2% and 5.2%), and the unidentified compound (RI 1778) (11.0% and 10.3%). The rhizome essential oil composition of *C. aromatica* show wide variation depending on geographical location [14]. For example, camphor was found to be a major component of *C. aromatica* rhizome essential oils from India (18.8–32.3%), whereas 8,9-dehydro-9- formylcycloisolongifolene (2.7–36.8%) was a dominant compound in the essential oils from China. Camphor was relatively minor in *C. aromatica* cultivated in North Alabama in this work (1.4–2.5%) and 8,9-dehydro-9-formylcycloisolongifolene was not observed. Curzerenone, germacrone, and 1,8-cineole, however, are relatively concentrated in Indian *C. aromatica* rhizome essential oils [14]. The rhizome essential oil of *C. aromatica* from Thailand showed camphor (26.9%), *ar*-curcumene (23.2%), and xanthorrhizol (18.7%) as the major components [30], whereas the rhizome essential oil from *C. aromatica* cultivated in Japan revealed β-turmerone (32.2 and 44.0%), 1,8-cineole (7.5 and 25.3%), and germacrone (4.6 and 9.6%) to be major compounds [31]. Notably, curzerenone, 8,9-dehydro-9-formylcycloisolongifolene, *ar*-curcumene, and xanthorrhizol were not detected in the essential oils from Japan. A recent examination of *C. aromatica* from different regions of eastern and southern India revealed relatively low concentrations of curzerenone (0.0–1.2%), but high concentrations of xanthorrhizol (8.8–24.4%), camphor (4.1–18.1%), germacrone (3.5–21.9%), neocurdione (5.8–14.6%), and 1,8-cineole (3.7–11.9%) [32].

The black-rhizome (*C. caesia*, CC38) essential oil, on the other hand, was rich in curzerenone (26.1% and 29.1%), curdione (28.7% and 35.6%), as well as *iso*-curcumenol (6.5% and 5.6%), for the mother and daughter rhizomes, respectively. In contrast, *C. caesia* rhizome essential oil from north India showed 8,9-dehydro-9-formylcycloisolongifolene, (11.7%), camphor (6.1%), 1,8-cineole (6.0%), and β-elemene (5.2%, reported as β-germacrene) as major components [33]. Neither curzerenone, curdione, nor *iso*-curcumenol were reported in the essential oil from north India, and 8,9-dehydro-9-formylcycloisolongifolene was not found in the essential oil from North Alabama. Note that curzerenone was determined to be artificially elevated in *C. caesia* essential oil due to the Cope rearrangemen<sup>t</sup> of furanodienone [34].




1459 1458 *allo*-Aromadendrene tr 0.1 tr 0.1 — — tr tr ———— 1462 1462 α-Acoradiene — — — — — — — — — — tr tr 1472 1474 Selina-4,11-diene — — — tr tr tr — ————— 1473 1473 Dauca-5,8-diene — — 0.1 0.3 tr tr 0.1 0.2 — — — — 1474 1475 γ-Muurolene 0.1 0.3 — — tr tr — —————

**CL63 (D)**





Figure S2. The major components are highlighted in **bold font**.

Likewise, the lime-colored rhizome essential oils of *C. zanthorrhiza* (CZ44,) were also rich in curzerenone (16.3% and 19.7%) and curdione (19.8% and 17.7%), in addition to germacrone (11.3% and 11.1%) for the mother and daughter rhizomes, respectively. This composition is in marked contrast to that reported for *C. zanthorrhiza* from Thailand with 1,8-cineole (37.6%) and curzerenone (13.7%) as the major components [30]. 1,8-Cineole was in lower concentration (7.2% and 2.5% for the mother and daughter rhizomes) in *C. zanthorrhiza* in this work. A *C. zanthorrhiza* rhizome essential oils from Bogor, Indonesia, on the other hand, was dominated by xanthorrhizol (26.8%), β-curcumene (17.0%), *ar*-curcumene (15.1%), camphor (9.1%), and germacrone (5.4%) [35]. Another sample of *C. zanthorrhiza* from West Java, Indonesia, was composed of β-curcumene (23.4%), *ar*-curcumene (22.1%), curzerene (6.0%), camphor (5.0%), and xanthorrhizol (4.7%) as the dominant constituents [36]. Neither β-curcumene nor xanthorrhizol were detected in the essential oil sample cultivated in North Alabama.

The major components of the rhizome essential oils from *C. longa* CL56 (yellowcolored rhizome) were α-turmerone (12.7% and 14.1%), α-zingiberene (11.4% and 13.9%), *ar*-turmerone (8.3% and 12.6%), and β-sesquiphellandrene (8.9% and 10.0%). The redcolored rhizome variety of *C. longa* (CL63) also yielded essential oils rich in *ar*-turmerone (36.1% and 31.3%), and α-turmerone (15.2% and 13.0%), as well as β-turmerone (=curlone) (15.4% and 13.0%). By comparison, the rhizome essential oils of *C. longa* cultivated in North Alabama, reported previously, showed α-turmerone (13.6–31.5%), *ar*-turmerone (6.8–32.5%), β-turmerone (4.8–18.4%), α-phellandrene (3.7–11.8%), 1,8-cineole (2.6–11.7%), α-zingiberene (0.9–12.5%), and β-sesquiphellandrene (0.7–8.0%) [9]. Two distinct chemical variations were found in the previous examination of *C. longa* cultivated in North Alabama. One group was dominated by turmerones ( α-turmerone, *ar*-turmerone, and β-turmerone), while the second group had lower concentrations of turmerones but high concentrations of α-zingiberene and β-phellandrene. Thus, the red-colored rhizome (CL63) belongs to the turmerone-rich chemical group, while the yellow-colored rhizome (CL56) belongs to the second group (high in α-zingiberene).

In order to place the volatile phytochemistry of the *Curcuma* rhizome essential oils in this study into perspective, an agglomerative hierarchical cluster analysis (HCA) was carried out based on the relative concentrations of the major components (Figure 2). There are two clearly defined clusters with at least 50% similarity based on the HCA: Cluster 1 is a cluster made up of CA22 (green rhizome), CA46 (*C. aromatica*, white rhizome) CC38 (*C. caesia*, black rhizome), and CZ44 (*C. zanthorrhiza*, lime rhizome), essential oils and defined by relatively high concentrations of curzerenone (14.7–29.1%), curdione (1.3–35.6%), and germacrone (3.8–14.7%); and Cluster 2, a cluster of CL56 (*C. longa*, yellow rhizome) and CL63 (*C. longa*, red rhizome) rhizome essential oils that were dominated by *ar*-turmerone (8.3–36.1%), α-turmerone (12.7–15.2%), and β-turmerone (5.0–15.4%) (see Table 2).

Interestingly, the volatile phytochemistry of *C. caesia* and *C. zanthorrhiza* rhizomes are very similar (about 90% similarity). Likewise, the green- and white-colored rhizome essential oils of *C. aromatica* are very similar (about 95% similarity). The yellow- and red-colored rhizome essential oils of *C. longa* showed somewhat lower similarity (about 60% similarity).

There are some significant differences in the concentrations of the major components in Cluster 1 (Figure 3). The concentration of curzerenone is significantly greater in *C. caesia* than in either *C. aromatica* or *C. zanthorrhiza*. The concentrations of curdione in *C. aromatica* are significantly lower than those in either *C. caesia* or *C. zanthorrhiza*. Germacrone was significantly lower in *C. caesia* than in either *C. aromatica* or *C. zanthorrhiza*.

**Figure 2.** Dendrogram obtained from the agglomerative hierarchical cluster analysis of the rhizome essential oil compositions of Vietnamese *Curcuma* species cultivated in North Alabama.

**Figure 3.** Comparison of the main chemical components of *Curcuma aromatica*, *Curcuma caesia*, and *Curcuma zanthorrhiza*. For each chemical component, bars with the same letter are not significantly different at *p* ≤ 0.05.

The significant differences between the essential oils of yellow- and red-colored *C. longa* are the concentrations of *ar*-turmerone (much higher in the red rhizome variety) and β-turmerone (also higher in the red rhizome variety). The concentrations of α-turmerone in the red and yellow varieties are not significantly different (Figure 4). Nevertheless, although the compositions of yellow- and red-colored rhizomes of *C. longa* are notably different (60% similarity), they are comparable to the respective chemical profiles of *C. longa* from tropical Asian collections [3].

**Figure 4.** Comparison of the main chemical components of *Curcuma longa*. For each chemical component, bars with the same letter are not significantly different at *p* ≤ 0.05.

#### *3.2. Enantiomeric Distribution of Terpenoids in Curcuma Essential Oils*

The enantiomeric distributions of terpenoid components in *Curcuma* rhizome essential oils have been determined by enantioselective GC-MS (Table 3). Although only found in trace quantities, when detected by chiral GC-MS (*C. aromatica*, *C. caesia*), the (−)-α-thujene predominated. (−)-α-Pinene was the dominant enantiomer in all samples. (−)-β-Pinene also predominated in all samples, but was especially dominant in the essential oils of *C. zanthorrhiza* (CZ44) and *C. longa* (CL56 and CL 63). (+)-Camphene was the dominant enantiomer in *C. aromatica*, *C. caesia*, and *C. zanthorrhiza*.

α-Phellandrene and δ-3-carene, only detected in the essential oils of *C. longa*, were both exclusively the (+) enantiomers. (−)-Limonene was the major stereoisomer in the *Curcuma* essential oils, but was nearly racemic in *C. zanthorrhiza*. (−)-Sabinene predominated is all *Curcuma* rhizome essential oils where it was detected. (−)-Linalool was the major enantiomer in nearly all *Curcuma* samples, but was particularly abundant in *C. caesia* (CC38), *C. zanthorrhiza* (CZ44), and the yellow-rhizome *C. longa* (CL56). Interestingly, however, the red-rhizome *C. longa* (CL63) exhibited (+)-linalool as the major enantiomer. Likewise, (−)- α-terpineol was the dominant enantiomer in all *Curcuma* essential oils. Camphor was not found in *C. longa*, but (+)-camphor was the dominant enantiomer in *C. aromatica*, *C. caesia*, and *C. zanthorrhiza*. The major enantiomer of terpinen-4-ol in *Curcuma* essential oils was (−)-terpinen-4-ol, although the distribution was nearly racemic in *C. zanthorrhiza*.

δ-Elemene was nearly racemic in all of the *Curcuma* essential oils, whereas *trans*-βelemene was exclusively the (−) enantiomer. Both (*E*)-β-caryophyllene and δ-cadinene were 100% (−) enantiomers, while germacrene D and β-bisabolene were exclusively the dextrorotatory stereoisomers.


**Table 3.** Enantiomeric distribution of terpenoid components in Vietnamese *Curcuma* rhizome essential oils cultivated in North Alabama.



(+) = dextrorotatory enantiomer, (−) = levorotatory enantiomer.
