Abstract
70 species of grasses family (Poaceae), coming from genera: Agrostis, Alopecurus, Anthoxanthum, Apera, Arrhenatherum, Avena, Brachypodium, Briza, Bromus, Calamagrostis, Corynephorus, Cynosurus, Dactylis, Danthonia, Deschampsia, Digitaria, Echinochloa, Elymus, Eragrostis, Festuca, Glyceria, Helictotrichon, Hierochloe, Holcus, Hordeum, Koeleria, Leymus, Lolium, Milium, Molinia, Nardus, Panicum, Phalaris, Phleum, Phragmites, Poa, Saccharum and Setaria, collected mostly from natural stands in Poland during 2020 season, were subjected to GC-MS fingerprinting of headspace volatile fraction above dried material. Obtained mass spectrometry data were analyzed by means of principal component analysis (PCA) and hierarchical cluster analysis (HCA). Five species: Glyceria maxima (Hartm.) Holmb., Lolium multiflorum Lam., Hordeum jubatum L., Bromus tectorum L. and Bromus secalinus L. were identified as outliers, which is consistent with our earlier analysis by thin layer chromatography. These species deserve further look and their outliance is orthogonal to coumarin content, which was independently observed for odorant species of grasses.
1 Introduction
Grasses (Poaceae) are a very important plant family, both in ecology and in the development of human civilization. They make up most cultivated crops and they are an important source of food (cereals), sugar (sugarcane), spices (lemongrass), as well as construction material (bamboo group). There is no phytochemical review covering whole family so far, but according to more specialized reviews [1–7], regardless of nutritional constituents they are rich source of biologically active substances.
Headspace gas chromatography is a broad and common technique of analysis of various types of samples [8]. The main advantage of this approach is that samples do not require (or require in a very limited range) an extraction or other pretreatment step. In case of plant material, fresh or dried plant can be placed directly in the sample vial and the gas phase above the sample is analyzed. The only disadvantage which is worth to mention is a limitation of the analysis to volatile compounds. However, due to a possibility of sample heating, a large part of biologically active phytochemical constituents can be detected with this method.
The comparative analysis of samples with GC-MS requires processing a large amount of data. One of available approaches is to deconvolute samples to a set of peaks and corresponding mass spectra [9]. Their intensities are then used for comparative analysis. This method works fairly well in metabolomics, but its disadvantage is that it is time, energy and resource consuming. Deconvolution of hundreds of samples can take hours and requires in practice to leave working computer overnight with full load of CPU. In classical chemometric approaches chromatograms are converted to matrices treated as “fingerprints” [10]. Regarding GC-MS samples, this results with huge amount of data. For instance, the matrix of one sample can consist of 1,500 time points (25 min of run with one second resolution) and 5,000 mass points (range 0–500 m/z with 0.1 resolution). Such large matrix, represented with standard double precision floating point arithmetics (8 bytes per one cell) occupies almost 60 MB of memory. Explorative analysis of 100 samples (6 GB) dataset with conventional unsupervised techniques can be done with at least 32 GB RAM, whereas hundreds or thousands of samples increase the required resources beyond current possibilities.
A matrix of GC-MS data can be converted to a TIC (Total Ion Count) vector by summing all masses for each time point. This process yields an univariate chromatogram, which occupies a small amount of memory and can be used to comparative analysis. However, the differences cannot be interpreted in context of mass spectra, as all MS information is lost.
A complementary approach is to sum chromatograms along time for each m/z value. The result is the average mass spectrum of a sample, almost equivalent to direct injection of a sample to mass detector [11]. This allows to a quick comparison and finding interesting masses, but does not provide information about a peak localization, nor the number of the peaks containing a particular mass.
An alternative to these two approaches is to combine these two matrices and analyze them simultaneously. As unsupervised methods search for the direction of largest variance, they should find mutual combinations of TIC and MS data, resulting in directions representing one or several peaks with masses attached to them. These two matrices can be treated as “data blocks” representing two datasets of various nature, characterizing the same samples. Recently, the combining of various datasets gained increasing popularity and such combined analysis is called “multiblock analysis” [12]. Dealing with these data raises new important challenges to be addressed, from proper preprocessing to careful selection of the explorative technique [13].
Multiblock analysis is frequently used in literature to combine data of different nature (signals from various detectors, frequently together with other non-signal results) and there are approaches allowing the holistic look across many datasets [14]. According to our knowledge, there were no trials to use this technique to combine TIC and MS signals into one analysis, to reduce substantially the amount of analyzed data allowing quick exploratory analysis before going deeper into the subject. Therefore we decided to check the usability of this approach on headspace GC-MS data of grasses.
Headspace technique is not very common in analysis of plants from grass family. Literature survey revealed that this technique was used so far for food products: rice [15, 16], sugarcane [17], oat [18, 19], oat flakes [20], proso millets [21], roasted barley tea [22], as well as for explorative profiling of wild plants: mixed hay identification [23], Festuca arundinacea Schreb. [24], Capillipedium parviflorum (R.Br.) Stapf [25], Cymbopogon schoenanthus (L.) Spreng. [26], Cymbopogon flexuosus (Nees ex Steud.) W.Watson [27] or Cynodon dactylon (L.) Pers. and Panicum repens L. [28]. There are also studies of fungi [29] and ozone [30] impact on volatile fraction of barley. Other studies are related to lactobacillus fermentation of grasses [31–33], aroma of bamboo [34, 35] and volatile environment pollutants in reed [36]. Due to lack of literature related to many grass species analyzed comparatively in one project, we decided to choose them as a source of data in this paper.
2 Experimental
2.1 Chromatographic conditions
The analyses of herbal samples were performed by a 7890A gas chromatography system coupled to a 7000 MS/Triple Quad mass spectrometric detector and combined with 7697A headspace sampler (Agilent Technologies, Palo Alto, USA). Separation was carried out using chromatographic capillary column HP-5MS ((5% phenyl-methylpolysiloxane), 30 m × 0.25 mm i.d., 0.25 µm film thickness) with helium of high purity (99.9999%, Messer, Chorzów, Poland) as a carrier gas at a constant flow 1.0 mL min−1. The mass spectrometer was tuned using perfluorotributylamine with masses m/z 69.0, 264.0 and 502.0. The GC column was operated in temperature-programmed mode with an initial oven temperature of 40 °C (held for 2 min), ramp at 250 °C with a 15 °C min−1 rate, and held at this temperature for 4 min. The temperature of injector, ion source, MS transfer line an quadrupoles were as follows: 250 °C, 230 °C, 300 °C and 150 °C, respectively. The Injections (1 μL) were done in the split mode with the split ratio 10:1. The mass detector was operated in scan mode and standard electron impact conditions were set (70 eV). To eliminate the metastable helium species, helium gas (2.25 mL min−1) was used as a quench gas. The data was carries out over a mass range of m/z from 30 to 500 at a rate of 4 scans s−1. The headspace sampler was connected to the GC system front inlet via a heated fused silica transfer line. A 1 ml sample loop was employed. The temperatures of headspace oven, loop and transfer line were set at 90 °C, 100 °C and 115 °C, respectively. The Injections (sampling time 0.8 min) were done in the flow to pressure mode (15psi). The system was operated by Agilent MassHunter B.07 (build 7, service pack 2) software. The extraction of organic compounds from the examined herbal samples was performed using 20 ml headspace vials containing 2 g of dried and ground material. The vials were sealed with PTFE-lined septum and an aluminium crimp cap, and then conditioned for 20 min at 90 °C. Once equilibrium was reached, the vials were pressurized to 15 psi within 1 min. Chromatograms were recorded 17 min, starting from 3rd minute of running (3–20 min).
2.2 Plant material and sample preparation
Panicum virgatum ‘Shanendoah’ plant material was kindly donated by Meadowlark Botanical Garden (Vienna, Virginia, USA). Saccharum officinarum was cultivated in Botanical Garden of Maria Curie-Skłodowska University in Lublin, Poland. All other material were collected by Joanna Wróbel-Szkolak during 2020 season in Poland from natural stands and habitats. Species were determined according to current literature by Anna Cwener. Herbarium specimens containing dates, locations and further information are available for the future reference. Table 1 contains the Latin names of the investigated species together with the abbreviations used along the paper.
The investigated grass species with unified botanical names according to current nomenclature in Kew botanical garden database (worldfloraonline.org)
| No. | Species | Abbreviation |
| 1 | Agrostis canina L. | AgCan |
| 2 | Agrostis capillaris L. | AgCap |
| 3 | Agrostis gigantea Roth | AgGig |
| 4 | Agrostis stolonifera L. | AgSto |
| 5 | Agrostis vinealis Schreb. | AgVin |
| 6 | Alopecurus geniculatus L. | AlGen |
| 7 | Alopecurus pratensis L. | AlPra |
| 8 | Anthoxanthum aristatum Boiss. | AnAri |
| 9 | Anthoxanthum odoratum L. | AnOdo |
| 10 | Apera spica-venti (L.) P.Beauv. | ApSpi |
| 11 | Arrhenatherum elatius (L.) P.Beauv. ex J.Presl & C.Presl. | ArEla |
| 12 | Avena fatua L. | AvFat |
| 13 | Brachypodium pinnatum (L.) P.Beauv. | BrPin |
| 14 | Brachypodium sylvaticum (Huds.) P.Beauv. | BrSyl |
| 15 | Briza media L. | BrMed |
| 16 | Bromus arvensis L. | BrArv |
| 17 | Bromus carinatus Haczyk. & Arn. | BrCar |
| 18 | Bromus inermis Leyss. | BrIne |
| 19 | Bromus japonicus Thunb. | BrJap |
| 20 | Bromus racemosus L. | BrRac |
| 21 | Bromus secalinus L. | BrSec |
| 22 | Bromus tectorum L. | BrTec |
| 23 | Calamagrostis arundinacea (L.) Roth | CaAru |
| 24 | Calamagrostis epigejos (L.) Roth | CaEpi |
| 25 | Corynephorus canescens (L.) P.Beauv. | CoCan |
| 26 | Cynosurus cristatus L. | CyCri |
| 27 | Dactylis glomerata L. | DaGlo |
| 28 | Dactylis polygama Horv. | DaPol |
| 29 | Danthonia decumbens (L.) DC. | DaDec |
| 30 | Deschampsia cespitosa (L.) P.Beauv. | DeCes |
| 31 | Deschampsia flexuosa (L.) Trin. | DeFle |
| 32 | Digitaria sanguinalis (L.) Scop. | DiSan |
| 33 | Echinochloa crus-galli (L.) P.Beauv. | EcCru |
| 34 | Elymus hispidus (Opiz) Melderis | ElHis |
| 35 | Elymus repens (L.) Gould | ElRep |
| 36 | Eragrostis minor Host | ErMin |
| 37 | Festuca arundinacea Schreb. | FeAru |
| 38 | Festuca gigantea (L.) Vill. | FeGig |
| 39 | Festuca guestphalica Boenn. ex Rchb. | FeGue |
| 40 | Festuca nigrescens Lam. | FeNig |
| 41 | Festuca ovina L. | FeOvi |
| 42 | Festuca pratensis Huds. | FePra |
| 43 | Festuca rubra L. | FeRub |
| 44 | Glyceria fluitans (L.) R.Br. | GlFlu |
| 45 | Glyceria maxima (Hartm.) Holmb. | GlMax |
| 46 | Helictotrichon pubescens (Huds.) Schult. & Schult.f. | HePub |
| 47 | Hierochloe odorata (L.) P.Beauv. | HiOdo |
| 48 | Holcus lanatus L. | HoLan |
| 49 | Hordeum jubatum L. | HoJub |
| 50 | Koeleria glauca (Spreng.) DC. | KoGla |
| 51 | Leymus arenarius (L.) Hochst. | LeAre |
| 52 | Lolium multiflorum Lam. | LoMul |
| 53 | Lolium perenne L. | LoPer |
| 54 | Milium effusum L. | MiEff |
| 55 | Molinia caerulea (L.) Moench | MoCae |
| 56 | Nardus stricta L. | NaStr |
| 57 | Panicum miliaceum L. | PaMil |
| 58 | Panicum virgatum 'Shenandoah' | PaVir |
| 59 | Phalaris arundinacea L. | PhAru |
| 60 | Phleum pratense L. | PhPra |
| 61 | Phragmites australis (Cav.) Trin. ex Steud. | PhAus |
| 62 | Poa angustifolia L. | PoAng |
| 63 | Poa annua L. | PoAnn |
| 64 | Poa compressa L. | PoCom |
| 65 | Poa nemoralis L. | PoNem |
| 66 | Poa palustris L. | PoPal |
| 67 | Poa pratensis L. | PoPra |
| 68 | Poa trivialis L. | PoTri |
| 69 | Saccharum officinarum L. | SaOff |
| 70 | Setaria pumila (Poir.) Roem. & Schult. | SePum |
The plant material was dried under appropriate conditions (room temperature, humidity about 50%). The dried material of whole plant (which conforms to Herba term in pharmacopoeias) were crushed in a mechanical grinder. 2.5 g of carefully weighted material was extracted three times with 25 ml of methanol – acetone – water (3:1:1) mixture by 15 min of ultrasonication in temperature 35 °C. Each time the obtained extract was filtered on a Buchner funnel. The extracts were joined in a round-bottomed flask and evaporated to dryness in 35–40 °C in vacuum. Each dried extract was dissolved in 5 ml of methanol. 1 ml of the extract was transferred to the headspace vial and evaporated to dryness at room temperature under the nitrogen flow.
2.3 Data processing and chemometric analysis
The obtained chromatograms were processed in R 4.1.3 running under R Studio with “multiblock” package available in CRAN library. It was done by importing TIC chromatogram from CSV file, which can be found in each data folder, as well as by converting whole dataset to MZXML with ProteoWizard “mzconvert” tool, then by importing inside R with “readMzXmlData” package with 0.1 m/z resolution.
3 Results and discussion
The first trial was to put grinded grasses directly into the headspace sample vial. However, this method resulted in chromatogram with unacceptable sensitivity. Therefore, we decided to perform liquid extraction, evaporate the extract and dissolve in methanol just like in our previous TLC work [37]. Although there is a risk of the loss of most volatile compounds, each evaporated extract had very strong herbal odor and the corresponding chromatogram presented much more peaks.
The two obtained matrices had 70 rows (samples). M/z matrix had 4,201 columns (masses from 30.0 to 450.0 with step 0.1), whereas TIC matrix had 1,201 columns (time points from 4 to 14 min with 5 s step). They were analyzed together using one of the simplest multiblock methods: Simultaneous Component Analysis [38]. This method works like PCA but decomposes two matrices to common scores and separated loadings.
Our preliminary trials covered various multiblock techniques, not so similar to PCA, like Generalized Procrustes Analysis [39], Multiple Factor Analysis [40] or PCA-GCA [41]. They failed to extract any useful information or gave similar results to SCA but with extremely long computation. SCA gave best results in almost invisible time. Four components with interpretable information were identified, responsible for 71%, 17%, 3.4% and 2.7% of total variance, respectively. The remaining components contained mainly noise.
Inspecting the SCA scores on Fig. 1A, one can conclude that the first PC represents the difference between all investigated grasses and HiOdo, as well as AnOdo. The inspection of loadings tells us that this variance is connected mainly with one peak having retention time 11.38 min (Fig. 2A) and masses 145.9 and 118. This peak was identified in GC-MS software MassHunter using NIST library as coumarin, which conform to the fact that HiOdo and AnOdo are most odorant species (they deserved to be named “odorata” in Latin).
The scores of SCA analysis: (A) components 1 and 2, (B) components 3 and 4
Citation: Acta Chromatographica 36, 1; 10.1556/1326.2022.01099
The loadings of SCA analysis for time block: (A) the first component, (B) the second component, (C) the third and (D) the fourth component
Citation: Acta Chromatographica 36, 1; 10.1556/1326.2022.01099
The second SCA component is responsible for masses 31.1 and 44 (Fig. 3B) and the time profile of this trend reveals decaying baseline. This variance can be interpreted as the tail of methanol peak (31.1), which was present in the sample in trace amounts after evaporation. Mass 44 is connected with carbon dioxide content in gaseous phase above the sample, which is also tailing across the chromatogram. Five species: GlMax, LoMul, HoJub, BrSec and BrTec have highest values of this component. Although it is connected with trend independent of the phytochemical content, their outliance is similar to our previous study on these grasses, done by thin-layer chromatography [37]. Therefore, they truly deserve further investigation.
The loadings of SCA analysis for m/z block: (A) the first, (B) the second, (C) the third and (D) the fourth component
Citation: Acta Chromatographica 36, 1; 10.1556/1326.2022.01099
The third SCA component is modelling three peaks (Fig. 2C, retention times 8.4, 10.16 and 11.6) which were identified with NIST library as the main peaks of siloxane column bleeding with common m/z values characterizing this phenomenon, such as 206.9 and 280.9 (Fig. 3C).
The fourth SCA component is modelling nonspecific low masses (Fig. 3D), which are present in many peaks across chromatogram (Fig. 2D). It contains also the shape change of coumarin peak. This component contains all remaining information, the next contain only noise (not shown).
4 Conclusions
Our investigation proves that simple multiblock methods, such as SCA, can be used in introductory explorative analysis of GC-MS data with substantial reduction of needed resources, as each sample, being a large matrix, can be compressed to two vectors being TIC and average mass spectrum. This method can separate main sources of variance, regardless of the nature of this source (one peak, tailing baseline, column bleeding with several peaks etc.). It can tell about the data structure and is a great tool for the choice of a preprocessing method—for example one can know which masses or time ranges should be filtered during analysis of whole uncompressed data.
Conflict of interest
The fourth author, Łukasz Komsta is a member of the Editorial Board of the journal. Therefore, the submission was handled by a different member of the editorial board, and he did not take part in the review process in any capacity.
References
- 1.↑
Gagliano, J.; Anselmo-Moreira, F.; Sala-Carvalho, W. R.; Furlan, C. M. What is known about the medicinal potential of bamboo? Adv. Tradit Med. (ADTM) 2021. https://doi.org/10.1007/s13596-020-00536-5.
- 2.
Kumar, P.; Singh, S.; Sharma, A.; Kaur, G.; Kaur, R.; Singh, A. N.; Arundo Donax, L. An overview on its traditional and ethnomedicinal importance, phytochemistry, and pharmacological aspects. J. Herbmed Pharmacol. 2021, 10(3), 269–280. https://doi.org/10.34172/jhp.2021.31.
- 3.
Karami, S.; Yargholi, A.; Sadati Lamardi, S. N.; Soleymani, S.; Shirbeigi, L.; Rahimi*, R. A review of ethnopharmacology, phytochemistry and pharmacology of cymbopogon species. Res. J. Pharmacognosy 2021, 8(3), 83–112. https://doi.org/10.22127/rjp.2021.275223.1682.
- 4.
Ekpenyong, C. E.; Akpan, E.; Nyoh, A. Ethnopharmacology, phytochemistry, and biological activities of cymbopogon citratus (DC.) Stapf extracts. Chin. J. Nat. Medicines 2015, 13(5), 321–337. https://doi.org/10.1016/S1875-5364(15)30023-6.
- 5.
Oladeji, O. S.; Adelowo, F. E.; Ayodele, D. T.; Odelade, K. A. Phytochemistry and pharmacological activities of cymbopogon citratus: a review. Scientific Afr. 2019, 6, e00137. https://doi.org/10.1016/j.sciaf.2019.e00137.
- 6.
Singh, A.; Lal, U.; Mukhtar, H.; Singh, P.; Shah, G.; Dhawan, R. Phytochemical profile of sugarcane and its potential health aspects. Pharmacognosy Rev. 2015, 9(17), 45–54. https://doi.org/10.4103/0973-7847.156340.
- 7.
Gebashe, F.; Aremu, A. O.; Finnie, J. F.; Van Staden, J. Grasses in South African traditional medicine: a review of their biological activities and phytochemical content. South Afr. J. Bot. 2019, 122, 301–329. https://doi.org/10.1016/j.sajb.2018.10.012.
- 8.↑
Wang, Y.; McCaffrey, J.; Norwood, D. L. Recent advances in headspace gas chromatography. J. Liquid Chromatogr. Relat. Tech. 2008, 31(11–12), 1823–1851. https://doi.org/10.1080/10826070802129092.
- 9.↑
Tsugawa, H.; Cajka, T.; Kind, T.; Ma, Y.; Higgins, B.; Ikeda, K.; Kanazawa, M.; VanderGheynst, J.; Fiehn, O.; Arita, M. MS-DIAL: Data-independent MS/MS deconvolution for comprehensive metabolome analysis. Nat. Methods 2015, 12(6), 523–526. https://doi.org/10.1038/nmeth.3393.
- 10.↑
Komsta, Ł. Chemometrics in fingerprinting by means of thin layer chromatography. Chromatogr. Res. Int. 2012, 2012, 1–5. https://doi.org/10.1155/2012/893246.
- 11.↑
González-Domínguez, R.; Sayago, A.; Fernández-Recamales, Á. Direct infusion mass spectrometry for metabolomic phenotyping of diseases. Bioanalysis 2017, 9(1), 131–148. https://doi.org/10.4155/bio-2016-0202.
- 12.↑
Westerhuis, J. A.; Kourti, T.; MacGregor, J. F. Analysis of multiblock and hierarchical PCA and PLS models. J. Chemometrics 1998, 12(5), 301–321. https://doi.org/10.1002/(SICI)1099-128X(199809/10)12:5<301::AID-CEM515>3.0.CO;2-S.
- 13.↑
Campos, M. P.; Reis, M. S. Data preprocessing for multiblock modelling – a systematization with new methods. Chemometrics Intell. Lab. Syst. 2020, 199, 103959. https://doi.org/10.1016/j.chemolab.2020.103959.
- 14.↑
Tanabe, K.; Hayashi, C.; Katahira, T.; Sasaki, K.; Igami, K. Multiblock metabolomics: an approach to elucidate whole-body metabolism with multiblock principal component analysis. Comput. Struct. Biotechnol. J. 2021, 19, 1956–1965. https://doi.org/10.1016/j.csbj.2021.04.015.
- 15.↑
Chen, Y.; Zhang, M.; Zhang, Y.; Deng, Y.; Wei, Z.; Tang, X.; Liu, G.; Li, P. Effects of germination and extrusion on volatile flavor compounds in brown rice. Scientia Agricultura Sinica 2021, 54(1), 190–202. https://doi.org/10.3864/j.issn.0578-1752.2021.01.014.
- 16.↑
Hernandez, H. P.; Hsieh, T. C.-Y.; Smith, C. M.; Fischer, N. H. Foliage volatiles of two rice cultivars. Phytochemistry 1989, 28(11), 2959–2962. https://doi.org/10.1016/0031-9422(89)80261-4.
- 17.↑
Nonato, E. A.; Carazza, F.; Silva, F. C.; Carvalho, C. R.; De Cardeal, Z. L. A headspace solid-phase microextraction method for the determination of some secondary compounds of Brazilian sugar cane spirits by gas chromatography. J. Agric. Food Chem. 2001, 49(8), 3533–3539. https://doi.org/10.1021/jf000896r.
- 18.↑
Sides, A.; Robards, K.; Helliwell, S.; An, M. Changes in the volatile profile of oats induced by processing. J. Agric. Food Chem. 2001, 49(5), 2125–2130. https://doi.org/10.1021/jf0010127.
- 19.↑
Cognat, C.; Shepherd, T.; Verrall, S. R.; Stewart, D. Comparison of two headspace sampling techniques for the analysis of off-flavour volatiles from oat based products. Food Chem. 2012, 134(3), 1592–1600. https://doi.org/10.1016/j.foodchem.2012.02.119.
- 20.↑
Klensporf, D.; Jeleń, H. H. Effect of heat treatment on the flavor of oat flakes. J. Cereal Sci. 2008, 48(3), 656–661. https://doi.org/10.1016/j.jcs.2008.02.005.
- 21.↑
Yang, Q.; Zhang, W.; Li, J.; Feng, B. Differentiation of fatty acid, aminno acid, and volatile composition in waxy and non-waxy proso millet. Food Sci. Technol. (Brazil) 2022, 42. https://doi.org/10.1590/fst.58320.
- 22.↑
Tatsu, S.; Matsuo, Y.; Nakahara, K.; Hofmann, T.; Steinhaus, M. Key odorants in Japanese roasted barley tea (Mugi-Cha) - differences between roasted barley tea prepared from naked barley and roasted barley tea prepared from hulled barley. J. Agric. Food Chem. 2020, 68(9), 2728–2737. https://doi.org/10.1021/acs.jafc.9b08063.
- 23.↑
Valtiner, S. M.; Bonn, G. K.; Huck, C. W. Characterisation of different types of hay by solid-phase micro-extraction-gas chromatography mass spectrometry and multivariate data analysis. Phytochem. Anal. 2008, 19(4), 359–367. https://doi.org/10.1002/pca.1062.
- 24.↑
Mayland, H. F.; Flath, R. A.; Shewmaker, G. E. Volatiles from fresh and air-dried vegetative tissues of tall fescue (Festuca arundinacea Schreb.): relationship to cattle preference. J. Agric. Food Chem. 1997, 45(6), 2204–2210. https://doi.org/10.1021/jf9701796.
- 25.↑
Saini, R.; Jaitak, V.; Guleria, S.; Kaul, V. K.; Babu, G. D. K.; Singh, B.; Lal, B.; Singh, R. D. Comparison of headspace analysis of volatile constituents with GC-MS analysis of hydrodistilled and supercritical fluid extracted oil of Capillipedium parviflorum. J. Essent. Oil Res. 2012, 24(3), 315–320. https://doi.org/10.1080/10412905.2012.677141.
- 26.↑
Bellik, F.-Z.; Benkaci-Ali, F.; Alsafra, Z.; Eppe, G. Effect of different parameters on volatile composition of the different parts of cymbopogon schoenanthus L. Spreng (poaceae) extracted by headspace solid-phase microextraction and hydrodistillation. J. Essent. Oil-Bearing Plants 2021, 24(4), 841–862. https://doi.org/10.1080/0972060X.2021.1960203.
- 27.↑
Deshmukh, Y.; Yadav, V.; Nigam, N.; Yadav, A.; Khare, P. Quality of bio-oil by pyrolysis of distilled spent of cymbopogon flexuosus. J. Anal. Appl. Pyrolysis 2015, 115, 43–50. https://doi.org/10.1016/j.jaap.2015.07.003.
- 28.↑
Svenberg, L.; Emmer, Å. Chemical diversity between three graminoid plants found in western Kenya analyzed by headspace solid-phase microextraction gas chromatography–mass spectrometry (Hs-Spme-Gc-Ms). Plants 2021, 10(11). https://doi.org/10.3390/plants10112423.
- 29.↑
Fiers, M.; Lognay, G.; Fauconnier, M.-L.; Jijakli, M. H. Volatile compound-mediated interactions between barley and pathogenic fungi in the soil. PLoS One 2013, 8(6). https://doi.org/10.1371/journal.pone.0066805.
- 30.↑
Dong, X.; Sun, L.; Agarwal, M.; Maker, G.; Han, Y.; Yu, X.; Ren, Y. The effect of ozone treatment on metabolite profile of germinating barley. Foods 2022, 11(9). https://doi.org/10.3390/foods11091211.
- 31.↑
Ming, T.; Qiu, D.; Zhou, J.; Li, Y.; Zhang, C.; Zhang, D.; Su, X. Analysis of the deodorization and aroma during fermentation of grass carp by lactobacillus plantarum. J. Chin. Inst. Food Sci. Technol. 2017, 17(10), 202–210. https://doi.org/10.16429/j.1009-7848.2017.10.027.
- 32.
Chmelová, Š.; Tříska, J.; Růžičková, K.; Kalač, P. Determination of volatile compounds in grass and maize silages using SPME and GC-MS [Stanovení těkavých látek v. travních a. kukuřičných silážích mikroextrakcí, na pevné fázi a. plynovou, chromatografií s. hmotnostně-spektrometrickou detekcí]. Chemicke Listy 2008, 102(12), 1138–1144.
- 33.
Rivero, M. J.; Keim, J. P.; Balocchi, O. A.; Lee, M. R. F. In Vitro fermentation patterns and methane output of perennial ryegrass differing in water-soluble carbohydrate and nitrogen concentrations. Animals 2020, 10(6), 1–16. https://doi.org/10.3390/ani10061076.
- 34.↑
Fu, S.-G.; Yoon, Y.; Bazemore, R. Aroma-active components in fermented bamboo shoots. J. Agric. Food Chem. 2002, 50(3), 549–554. https://doi.org/10.1021/jf010883t.
- 35.↑
Zheng, J.; Zhang, F.; Zhou, C.; Lin, M.; Kan, J. Comparison of flavor compounds in fresh and pickled bamboo shoots by GC-MS and GC-olfactometry. Food Sci. Technol. Res. 2014, 20(1), 129–138. https://doi.org/10.3136/fstr.20.129.
- 36.↑
Mothes, F.; Reiche, N.; Fiedler, P.; Moeder, M.; Borsdorf, H. Capability of headspace based sample preparation methods for the determination of methyl tert-butyl ether and benzene in reed (Phragmites australis) from constructed wetlands. Chemosphere 2010, 80(4), 396–403. https://doi.org/10.1016/j.chemosphere.2010.04.024.
- 37.↑
Wróbel-Szkolak, J.; Cwener, A.; Komsta, Ł. Data fusion from several densitometric modes in fingerprinting of 70 grass species. JPC-J Planar Chromat 2022, 35(3), 287–297. https://doi.org/10.1007/s00764-022-00180-6.
- 38.↑
Levin, J. Simultaneous factor Analysis of several gramian matrices. Psychometrika 1966, 31(3), 413–419. https://doi.org/10.1007/BF02289472.
- 39.↑
Gower, J. C. Generalized Procrustes analysis. Psychometrika 1975, 40(1), 33–51. https://doi.org/10.1007/BF02291478.
- 40.↑
Pagès, J. Collection and analysis of perceived product inter-distances using Multiple factor Analysis: application to the study of 10 white wines from the Loire valley. Food Qual. Preference 2005, 16(7), 642–649. https://doi.org/10.1016/j.foodqual.2005.01.006.
- 41.↑
Smilde, A. K.; Måge, I.; Næs, T.; Hankemeier, T.; Lips, M. A.; Kiers, H. A. L.; Acar, E.; Bro, R. Common and distinct components in data fusion. J. Chemometrics 2017, 31(7), e2900. https://doi.org/10.1002/cem.2900.
