Investigation of fiber maturity and crystallinity information in Upland seed cottons by Fourier transform infrared spectroscopy

Abstract

Cotton fiber maturity has been determined by cross-sectional image analysis (IA), advanced fiber information system (AFIS), and Cottonscope methods on cotton lint. These methods have reported the results as average maturity and maturity distribution in a sample, through measuring the fibers in the ways of either individual fiber cross-section or longitude of several sub-samples. Previous studies have shown good agreement in maturity for well-prepared samples among these methods, although AFIS is observed to be less sensitive. As a different approach, attenuated total reflection Fourier transform infrared (ATR FT-IR) spectroscopy was proposed to measure fiber maturity (M_IR) at bundle fiber level. Extending fiber maturity measurement into seed cotton, the FT-IR method might be an option considering such factors as essential cotton seed and visible trash removal, measuring system availability and speed, and also sub-sampling representation in a naturally variable sample. A comparison of fiber M_IR average in seed cottons with AFIS maturity ratio (M_AFIS) in ginned cotton fibers exhibited a general trend of increasing M_AFIS with M_IR. On the basis of M_IR value, 3-M_IR (low-, mid-, and high-) fiber classification analysis implied the distinctions within seed cottons having close M_IR average, and among the same cultivar grown at different conditions. Additionally, cultivars with a similar maturity distribution varied in fiber crystallinity (CI_IR) distribution, and vice versa.

Keywords

Fourier transform infrared spectroscopy attenuated total reflection seed cotton fiber maturity maturity frequency crystallinity algorithm analysis

Cotton crop is one of the most important agricultural commodities in the world, primarily for its naturally produced textile fiber.¹ Cotton fiber is produced from protodermal cells on the outer integument layer of cotton seeds. It takes approximately 6 to 8 weeks for fiber to reach fully maturity. To meet the need of cotton fiber mechanical harvesting, chemical boll openers and leaf defoliants are used to increase cotton boll opening and reduce plant vegetative growth.² Maturity of fiber in seed cotton is naturally variable within locules in a single boll, from the bottom to the top boll position in a plant, and from different areas in a growth field, leading to fiber-to-fiber maturity variation within a sample.^3,4 Cotton fiber maturity is a key yield index and also an essential fiber quality attribute, which is directly linked to fiber breakage and entanglement (neps) during mechanical processing, dye uptake in yarn and fabric products, as well as yarn processing and textile performance.^5–9

All fiber maturity has been measured on cotton lint (i.e., cotton fiber collected during the ginning process to separate from the cotton seed), following established cotton fiber maturity testing procedures in the cotton industry. Current-in-use cotton fiber maturity measurements include the traditional cross-sectional image analysis (IA),^10–12 advanced fiber information system (AFIS),^13,14 and Cottonscope.^14–17 In addition, other techniques and systems have been attempted, for example, Fineness and Maturity Tester (FMT),¹⁸ polarized light microscopy,^19,20 confocal microscopy,⁴ near infrared (NIR), and Fourier transform infrared (FT-IR) spectroscopies.^21–23

Cross-sectional IA measurement was developed as a direct reference fiber maturity method in the 1980s,¹¹ and then improved continuously over the years.^24–28 It requires a bundle of fibers to be selected and combed, embedded in a medium, cut into thin sections, and mounted on a microscope slide, prior to capturing the images of fiber cross-sections using a microscope. Following image analysis with dedicated software, it is possible to calculate fiber characteristics, such as circularity (θ), maturity ratio, and fineness for each cross-sectional fiber as well as average maturity ratio (or maturity) and maturity distribution in a sample. The IA maturity (M_IA) readings have been utilized as fiber maturity references for indirect measurement system development, for instance, AFIS and Cottonscope. The IA system is comparatively inexpensive and is available at fiber research laboratories. However, the M_IA test is destructive, labor intensive, time-consuming, influenced by operators’ experience, and impractical for measuring a large number of fiber samples. In addition, over 300 cross-sectional fibers in an IA measurement were the analyses of fibers at a single fiber level, in which they were taken from several fiber bundles (or sub-samples) in a blended sample as fiber blending is part of the procedure for preparing fiber cross-sections. This type of fiber sub-sampling is acceptable and reasonable for homogenized and clean fibers such as the 104 reference materials,¹⁰ but it may not be suitable for regular or unblended cotton fibers due to their natural maturity variability for the purpose of maturity consistency and comparison from different measurements.^14,16,29,30

In the 1990 s, a rapid and automatic instrument known as AFIS was developed to measure fiber maturity and maturity distributions, and also fiber length and trash parameters. After a 30-cm-long sliver was prepared from a 0.5 g cotton fiber and fed into an AFIS system, the maturity and other fiber properties were obtained in less than two minutes as individualized fibers were moved by air through a light beam. AFIS maturity ratio (M_AFIS) was indirectly determined by measuring the shape of individualized fibers from two different angles. This test is relatively fast and might be less impacted by operators’ experience, but the procedure is destructive. For each measurement 3000∼5000 fibers were analyzed at single fiber level, and several measurements were performed on sub-samples within a sample. Like IA measurement, M_AFIS is acceptable for well-prepared fibers but may not be effective for regular cotton fibers when comparing M_AFIS values to those from Cottonscope and FT-IR measurements.^{14,15,29–31}

In 2010, Cottonscope was developed for determining fiber maturity, maturity distribution, fineness, and ribbon width simultaneously by using polarized microscopy. In the measurement, a fiber bundle was cut into snippets or sections around 0.7 mm in length first, then 50 mg of snippets were individualized in a water bowl with a stirring water solution to help the snippets' separation and distribution fast before capturing images of the weighed snippets (∼20,000). Like the AFIS test, the Cottonscope maturity (M_CS) test is rapid, the measurement is destructive, and the system is available at only a few fiber testing locations. Notably, 20,000 snippets were analyzed individually for each of several sub-samples in a sample. Similar to IA and AFIS method, Cottonscope is suitable for well-prepared fiber samples for maturity comparison.^14,15,30

Over the years, a number of studies have been undertaken to understand either how the M_AFIS and M_CS value agrees with M_IA or whether M_AFIS matches with M_CS.^{8,14–16,30,32} Through using 93 samples from an original set of the 104 well-defined cottons, Rodgers et al.¹⁶ compared fiber M_IA to M_CS and observed a good overall agreement between the two methods. Meanwhile, Rodgers et al.³⁰ reported a lower response of M_AFIS to 20 of the 104 reference materials than those of the M_IA and M_CS measurement, and also observed less sensitivity in the AFIS method of developing MD 90ne fiber than in the Cottonscope result. In a separate study of the same 104 reference samples, Paudel et al.¹⁴ concluded that both M_AFIS and M_CS measurements correlated well with the referenced cross-sectional M_IA method. Although M_AFIS correlated well with M_CS, M_AFIS was reported to be in a narrow range. The authors addressed the application of AFIS and Cottonscope to a small amount of fiber sample (at single boll level); however, they cautioned that it could be difficult to obtain a representative sample of the lint for the analysis. On a set consisting of 550 fiber samples from a multiparent advanced generation intercross (MAGIC) population, Kim et al.¹⁵ noted a difference in median maturity value between Cottonscope and AFIS measurements, and a weak correlation between the two maturities. Aiming to examine whether the fiber blending process impacts their maturity values, the authors blended 27 samples individually. Their results showed significant and positive correlations in both M_CS and M_AFIS between unblended and blended samples. Especially, they emphasized that homogeneity of cotton fibers influenced M_AFIS measurement more than M_CS measurement.

As a different approach, attenuated total reflection FT-IR (ATR FT-IR) spectroscopy was proposed in 2011,²⁹ based on spectral intensity differences induced by relative cellulose amount of cotton fibers during the growth (or the thickening of secondary cell wall). In the concept, two algorithms were created from fibers in seed cottons, and then infrared maturity (M_IR) index was compared with M_IA and M_AFIS on two sets of diverse cotton lint. Their results indicated a strong determination of correlation (R²) of 0.920 between M_AFIS vs. M_IR on six International Cotton Calibration (ICC) standards that were used to calibrate the high-volume instrument (HVI) systems in cotton industry, which differed from a much lower R² of 0.403 among 99 regular commercial cotton fibers and of 0.596 on 104 reference materials. The authors contributed low correlations between the pairs of M_AFIS vs. M_IR or M_IR vs. M_IA as the heterogeneous distribution of structural, physical, and chemical characteristics in maturity variable fibers and subsequent different sampling specimens. Recent examination on the developing and developed fibers validated maturity consistency and equivalency between the ATR FT-IR method and referenced cross-sectional IA method.³³ Simultaneously, two independent algorithms were proposed to estimate fiber crystallinity index (CI_IR) from the same spectra.^29,34 One ATR FT-IR spectrum represents the information of fibers at a bundle fiber level, differing from the analyses of fibers at individual fiber cross-section and longitude level in which fibers were aligned or cut prior to IA, AFIS, and Cottonscope test. Hence, conventional FT-IR measurement provides fiber maturity in general, while a large number of FT-IR sub-samplings across one sample is necessary for maturity distribution. During spectral collection, a small bundle of fibers (∼0.5 mg) was put into the optical window (2 mm in diameter) of an ATR device without any fiber pretreatment. Caution is needed when applying this method to estimate M_IR in less than 17 days of post-anthesis (DPAs) due to the effect of the physiological sugars on the two bands (1032 and 956 cm⁻¹) used in algorithm R₁.³⁵

There is a concern that ATR FT-IR measurement is a surface reflectance technique, since IR light is directed onto an internal reflection element (IRE) crystal with a high refractive index and excites the surface of the sample. The sampling depth of the ATR method was reported to be approximately 2 to 15 μm depending on ATR crystal materials and also to increase with decreasing wavenumber.^36–38 For instance, sampling depth increased from ∼1 μm for germanium crystal to over 6 μm for ZnSe crystal in the region of 1100 to 600 cm⁻¹.³⁶ Because the thickness of the secondary cell wall in mature cotton fibers varies from 2 to 7 μm,³⁹ the ATR method is capable of providing the information inside mature cotton fibers by the use of both low refractive index crystal (ZnSe or diamond) and low spectral region (1100 to 600 cm⁻¹), as is the case in this study and other fiber biosynthesis investigations.^40–42 Similar to IA, AFIS, and Cottonscope measurements, the FT-IR method has been performed at laboratory condition.

Furthermore, IR imaging spectroscopy has been attempted by Santiago et al.³¹ Their IR imaging system was equipped with a focal-plane array (FPA) detector and has the ability to cover a larger sampling area (15 × 15 mm) than a traditional ATR device. In this recent study, two of three IR band intensities (1032 and 956 cm⁻¹) identified previously²⁹ were used to analyze both imaging and ATR spectra at bundle fiber level. Their results indicated that average maturity values for a set of 30 samples showed a strong linearity and a R² value of 0.95 between the Cottonscope and FPA IR method, and also a compromised the relationship (R² = 0.45) between AFIS and FPA IR determination.

Extending fiber maturity measurement in cotton lint to seed cotton fiber may meet some difficulties, since seed cotton contains non-lint materials (or trash). From the aspect of measuring procedures, IA and FT-IR methods do not need to remove cotton seed and trash initially due to their micro-sampling advantage. With regard to sampling amount, all methods discussed above (apart from AFIS) are feasible, simply because IA, Cottonscope, and FT-IR require much less fiber (from several hundred single fibers to ∼50 mg) than AFIS (∼0.5 g). Fiber from one locule in one seed cotton boll could be sufficient to the fulfill maturity test by using the three methods, although cotton seeds need to be removed first prior to fiber cutting for the Cottonscope test. All fiber from one cotton boll could be enough for the AFIS test, but cotton seeds need to be separated first. Consideration should also be given to the fact that both IA and Cottonscope methods are relatively time-consuming and destructive, and also only represent the fiber maturity of several sub-samples in a total sample, which is reasonable if the fiber either is homogenous after a well-blending process at a large scale or is accessible in small amounts during cotton breeding and genetic studies. In other words, it could be a challenge to have a reliable fiber maturity of naturally variable seed cottons by several sub-samples from IA or Cottonscope measurements. Nevertheless, IA, AFIS, and Cottonscope provide maturity distribution that characterizes the level of maturity variation in a sample. The FT-IR method might be an alternative and complementary option for investigating fiber maturity in seed cottons, since it is capable of analyzing as little as 0.5 mg fiber directly without the need to remove any visible trash materials and cotton seeds or to prepare fiber samples (such as fiber alignment, cutting, and weighing) first, and also it is capable of collecting individual sub-samples (or locule) rapidly for less than 2 min. Because cotton lint was produced from a number of cotton boll locules, the lint is a mixture of fibers with varying maturity. By analyzing seed cottons at single locule level, it is possible to generate accurate maturity variation within a sample. The main objective of this study was to compare fiber M_IR average on 50 locules (or sub-samples) within each of 11 seed cotton samples to M_AFIS, and to classify individual locule fiber into one of three fiber classification classes on the basis of its M_IR and CI_IR values for qualitative comparison. The aim is to validate the usefulness and sensitivity of the FT-IR technique in monitoring the M_IR and CI_IR difference within one seed cotton sample or between seed cotton samples.

Experimental details

Seed cottons

A total of 11 commercial Upland seed cotton samples were collected at modules in the field randomly from three US states in 2019 crop year. Three cultivars (Deltapine 1646 (DP1646), Stoneville 4848 (ST4848), and Stoneville 5471 (ST5471)) were grown in Mississippi (MS), five samples from four cultivars (Deltapine 1646 (DP1646), DP1646 dryland (not irrigated), Phytogen 350 (PHY350), Phytogen 430 (PHY430), and Americot NG3522 (NG3522)) were from Missouri (MO), and three cultivars (Dyna-Gro 3385 (DG3385), Fiber Max 1830 (FM1830), and Phytogen 444 (PHY444)) were grown in New Mexico (NM). These samples were kept at laboratory environment (a constant temperature of 21°C and relative humidity of 65%) for 2 days before ATR FT-IR spectral acquisition.

ATR FT-IR collection, M_IR and CI_IR determination, and data analysis of seed cottons

An FTS 3000MX FTIR spectrometer (Varian Instruments, Randolph, MA) equipped with a ceramic source, KBr beam splitter, deuterated triglycine sulfate (DTGS) detector, and an ATR attachment was used to acquire the fiber spectra in seed cottons. The ATR sampling device utilized a DuraSamplIR single-pass diamond-coated internal reflection accessory (Smiths Detection, Danbury, CT), and a consistent contact pressure was applied by way of a stainless-steel rod and an electronic load display. For each cultivar, 50 locules were randomly selected across a sample box containing more than 10 lb of seed cottons, in which fibers were mingled together between locules and locules as well as between locule fibers and non-lint material. The ATR FT-IR spectrum was collected directly on each locule fiber over the range of 4000–600 cm⁻¹ at 4 cm⁻¹ resolution and 16 co-added scans, with extreme attention to avoid the spectral interference of any non-lint materials (seed, trash, seed-coat, etc.) in a locule. Hence, there are no effects of non-lint materials on M_IR or CI_IR calculations. No further spectral processing (such as baseline correction) was applied to these spectra in absorbance units, in identical steps to earlier ATR FT-IR investigation of cotton fibers for result consistency and comparison.

Fiber M_IR and CI_IR indices in seed cottons were calculated by performing simple algorithmic analyses of ATR FT-IR spectra as exampled in Figure 1 with reported procedure.^29,33,34 Briefly, after the spectra were exported into Microsoft® Excel® for Office 365, R₁ was calculated from the first algorithm R₁:

R_{1} = (I_{956} - I_{1500}) / (I_{1032} - I_{1500})

(1)

in which I₁₅₀₀, I₁₀₃₂, and I₉₅₆ are each a three-point intensity average at respective wavenumber. The wavenumber at 1032 cm⁻¹ represented the large and positive intensity variation, while that at 956 cm⁻¹ characterized the large and negative intensity variation in the difference spectrum between immature and mature cottons. The wavenumber at 1500 cm⁻¹ was used to offset two readings because of its minimum absorbance.

Next, the R₁ value was converted to M_IR by the second algorithm M_IR:

M_{IR} = (R_{1} - 0. 14) / 0. 45

(2)

Similarly, the CI_IR calculation consisted of two algorithms, with the first algorithm R₂ utilizing three respective IR intensities at 800, 730, and 708 cm⁻¹, and the second algorithm CI_IR (%) changing R₂ values into fiber CI_IR.

R_{2} = (I_{708} - I_{800}) / (I_{730} - I_{800})

(3)

C I_{IR} (%) = ((R_{2} - 1.4) / 2 .0) \times 100

(4)

There were 50 M_IR and CI_IR values for each of the 11 seed cotton samples. After classifying the M_IR or CI_IR in the respective range of 0.0 ∼ 1.0 or 0.0 ∼ 100.0 into 20 bins (bin width, 0.05 or 5.0) and counting the frequency number of each bin, a M_IR or CI_IR frequency distribution curve was obtained for each individual sample. Separately, p-values at confidence level of 95% were calculated using regression function under Data Analysis in Microsoft® Excel® for Office 365.

Seed cotton ginning, fiber M_AFIS and M_IR measurement in ginned lint

Seed cottons were ginned by using a laboratory roller gin. The ginned fibers were conditioned at 21 ± 1°C temperature and 65 ± 2% relative humidity for 48 hours prior to AFIS testing. The M_AFIS value was determined by an AFIS Pro 2 (USTER Technologies Inc., Knoxville, TN) with three replications of 5000 fibers per measurement. Separately, 50 sub-samples were randomly taken across each ginned fiber sample and their FT-IR spectral analyses were identical to the seed cottons described above.

Results and discussion

Comparison of M_IR on 11 seed cotton samples, and M_AFIS in ginned samples

Table 1 summarizes the M_IR average, standard deviation (SD), and range of 11 seed cotton samples, showing a variation of M_IR from 0.676 to 0.900 in average and from 0.323 to 1.000 at single locule level. Compared to the least mature (M_IR = 0.0) and most mature (M_IR = 1.0) samples in the original dataset,²⁹ none of the locules in this study was considered as the least mature, but some locules from eight different samples were considered to be the most mature. There were significant differences (p-value <0.001) of 50 locule fibers between any two of 11 samples. In general, SD decreased linearly and significantly with the M_IR average increasing (Figure 2), indicating a larger variation of M_IR value within lower M_IR fibers than within higher M_IR fibers. This observation implied the effect of cotton genotype, environment (weather, location, and irrigation), and their interactions on fiber maturity.

Table 1.

Comparison of M_IR ave ± SD and range on 11 seed cotton samples, and also mean M_IR and M_AFIS in ginned samples. Difference in mean M_IR between seed cotton and ginned fibers was given in parenthesis

Location/cultivar	Seed cottons		Ginned cottons (ave)
Location/cultivar	M_IR ave ± SD	M_IR range	M_IR ave ± SD(Seed-Ginned)	M _AFIS
MS
DP1646	0.844 ± 0.103	0.531∼1.000	0.808 ± 0.114 (0.036)	0.91
ST4848	0.864 ± 0.127	0.482∼1.000	0.807 ± 0.160 (0.057)	0.91
ST5471	0.736 ± 0.149	0.441∼1.000	0.813 ± 0.121 (−0.077)	0.94
MO
DP1646	0.815 ± 0.121	0.440∼0.983	0.755 ± 0.158 (0.060)	0.89
DP1646 dryland	0.800 ± 0.106	0.497∼1.000	0.780 ± 0.126 (0.020)	0.83
NG3522	0.801 ± 0.115	0.513∼0.993	0.818 ± 0.108 (−0.017)	0.90
PHY350	0.866 ± 0.099	0.553∼1.000	0.795 ± 0.116 (0.071)	0.90
PHY430	0.851 ± 0.108	0.485∼1.000	0.745 ± 0.131 (0.106)	0.88
NM
DG3385	0.676 ± 0.164	0.323∼0.944	0.725 ± 0.141 (−0.049)	0.80
FM1830	0.900 ± 0.077	0.637∼1.000	0.848 ± 0.090 (0.052)	0.91
PHY444	0.742 ± 0.150	0.352∼1.000	0.718 ± 0.145 (0.024)	0.85

As expected in Table 1, samples with similar M_IR average could show different maturity ranges. For example, the pair of ST4848 vs. PHY350 revealed a similar M_IR average of 0.864 vs. 0.866 but a different M_IR range of 0.482 to 1.000 and 0.553 to 1.000, respectively, and also the pair of ST5471 vs. PHY444 had a close M_IR value (0.736 vs. 0.742) but a respective M_IR range of 0.441 to 1.000 and 0.352 to 1.000. On the other hand, cultivars with different M_IR average could have the similar M_IR range, as the case of ST5471 cultivar against DP1646 cultivar grown in MO, in which both had a respective M_IR average of 0.736 and 0.815 but shared a close M_IR range of 0.441 to 1.000 vs. 0.440 to 0.983. Comparatively, DP1646 dryland and NG3522 showed a common M_IR mean (0.800 vs. 0.801) and M_IR range (0.497 to 1.000 vs. 0.513 to 0.993), and ST4848 cultivar resembled PHY430 cultivar in either M_IR average (0.864 and 0.851) or M_IR range (0.482 to 1.000 vs. 0.485 to 1.000).

For DP1646 cultivar, the fiber sample from MS had a greater M_IR average than those from MO (0.844 vs. 0.800 ∼ 0.815), with the smallest M_IR average (0.800) for dryland fibers, while two PHY cultivars (PHY350 vs. PHY430) from MO did not show a difference in M_IR average (0.866 vs. 0.851).

In the preceding study,²⁹ Liu et al. observed that R₁ readings were in the range of 0.35–0.58 for 201 mature fibers and 0.16–0.41 for 201 immature samples. With an R₁ threshold value at 0.40, 98.0% immature fibers and 94.5% mature samples were correctly classified. Further, they represented the R₁ values as M_IR by an additional algorithm. Thus, immature fibers whose R₁ < 0.40 corresponded to a M_IR < 0.58 in the maturity range of 0 to 1.0, and vice versa. Further examination of published Figure 3 in the citation²⁹ showed that there existed a turning point around the R₁ = 0.50 that corresponded to M_IR = 0.80. Hence in this study, fiber samples with M_IR average of less than 0.58, 0.58 to 0.80, or greater than 0.80 are subjectively classified as one of 3-M_IR classes, i.e., low-maturity (or less mature), mid-maturity (or mature), or high-maturity (more mature) fiber. Under this M_IR scale, none of 11 seed cotton samples as a whole was rated to low-maturity fiber. In turn, a total of seven samples (DP1646 and ST4848 from MS, DP1646, NG3522, PHY350, and PHY430 from MO as well as FM1830 from NM) was classified as high-maturity fibers, whereas the remaining four samples were considered as mid-maturity fibers.

Figure 3 relates the fiber M_IR averages on seed cottons to M_AFIS on ginned cottons. Mean M_IR values in seed cottons varied from 0.676 to 0.900, whereas M_AFIS in ginned samples changed from 0.80 to 0.94. With 11 Upland cotton samples showing a narrow M_AFIS range, there was a general pattern of M_AFIS increasing with M_IR rising (R² = 0.30), and also there was a non-significant difference statistically between M_AFIS and M_IR value (p-value = 0.08). Notably, there was a sample (ST5471) that showed relative low M_IR value (0.736) but high M_AFIS (0.94). To compare M_IR values in seed cottons, FT-IR spectra of 50 sub-samples in their ginned samples were collected randomly across individual sample and resultant M_IR were inserted in Table 1. Difference in mean M_IR between seed cottons and their ginned fibers was not statistically significant (p-value = 0.06). In general, the differences, ranging from −0.077 to 0.106, were small and acceptable for all samples including the ST5471 sample. Without the ST5471 sample, the remaining 10 seed cottons showed an improved relationship between M_IR and M_AFIS (R² = 0.74), suggesting that a minimum selection of 50 randomly locules in current work is acceptable. Meanwhile, all 11 seed cottons were analyzed during the same time by an identical procedure in this investigation. In general, a R² of 0.30 in Figure 3, between M_AFIS on ginned fiber and M_IR on seed cotton fiber, is lower than reported R² values between M_AFIS and M_IR measured on ginned fiber; for example, a 0.920 for 6 HVI micronaire (MIC) calibration standards (2.52 to 5.35 MIC range) that each sample was well-blended and pre-selected for consistency,²⁹ a 0.45 for 27 fiber samples including cotton MIC standards, developing and developed Upland fibers,³¹ and a 0.403 for 99 commercial cotton fibers from diversified sources.²⁹ Even tested on ginned fiber, the comparison between M_AFIS and M_CS measurements has been inconsistent, such as a good agreement for partial or all of the 104 well-blended samples representing a wide range of fiber maturity and different cotton varieties^14,30 but a weak correlation for 550 Upland MAGIC fiber samples from the breeding scheme.¹⁵ Although it showed a relatively low R² (=0.30) in Figure 3, the difference between M_AFIS and M_IR values was not significant statistically (p-value = 0.08). Nevertheless, these comprehensive results imply the difficulty of linking M_AFIS value with M_CS or M_IR for unblended fibers in lint,^15,29 and further a challenge of relating M_AFIS value from ginned lint to M_IR value from naturally variable seed cottons. The barriers include fiber maturity variability, less sensitivity of AFIS maturity measurement, and substantial difference in measuring fiber maturity between two methods (M_IR: directly measures the amount of cellulose and therefore the maturity at bundle fiber level vs. M_AFIS: indirectly determines fiber maturity from the shape of individualized fibers at single fiber level).

Comparison of M_IR frequency on 11 seed cotton samples

Figures 4(a) to (c) provide M_IR frequency of locules on seed cottons from three grown locations (MS, MO, and NM), respectively. They exhibited a broad range of M_IR variation within a seed cotton cultivar (0.3 to 1.0) and showed different frequency patterns among one growing location or between three locations. This indicated that cotton genotype and environment (weather, location, and irrigation), as well as their interactions, greatly influence fiber M_IR property. Opposite to one cultivar (ST5471) from MS with a maximum M_IR frequency of 20% occurred at approximately lower M_IR = 0.70, the other two cultivars (DP1646 and ST4848) from the same location showed a similar M_IR frequency distribution with a maximum M_IR frequency of 25% or more at higher M_IR = 0.85. Among five seed cotton samples from MO, four of them (DP1646, NG3522, PHY350, and PHY430) indicated a great M_IR frequency occurrence at higher M_IR = 0.85 and 0.95, whereas the remaining one (DP1646 dryland) had large M_IR occurrence at lower M_IR = 0.75. Three seed cotton cultivars from NM suggested a clear M_IR frequency discrepancy between two cultivars (DG3385 and PHY444) and another one (FM1830), with the former two cultivars showing an unclear maximum and less than 20% M_IR frequency across entire M_IR range and the latter one cultivar indicating a maximum M_IR frequency of 25% and more at higher M_IR range (0.9 ∼ 1.0). Overall, on 11 seed cotton samples examined in this study, two cultivars (DG3385 and PHY444) showed less than 20% M_IR frequency across entire M_IR range, while other cultivars indicated greater than 20% M_IR frequency occurrence at several and narrow M_IR intervals.

None of the M_IR frequency patterns in Figure 4 appeared to be a normal Gaussian shape, compared to either M_AFIS distribution of ginned fibers in Figure 5 or earlier reports from the IA, the Cottonscope, as well as ATR and FPA maturity distribution of either well-blended fibers or the developing fibers.^{8,25,30,31,43} Figure 5 showed a Gaussian M_AFIS distribution, in which 5000 fibers in one AFIS measurement were analyzed and the results were reported on three bundle fiber sub-samples (∼15,000 fibers). Comparatively, over 200, cross-sectional fibers in each M_A test were the analyses of fibers at an individual fiber cross-section level, and average results were obtained by measuring several bundle fiber sub-samples in a total sample. In addition, 20,000 snippets in one M_CS measurement were the analyses of cut fibers at an individual fiber level longitudinally, and results were acquired by testing three or more bundle fiber sub-sample cuttings. Both 144 datapoints or spectra (36 spectra/replicate × 4 replicates) in FPA and 150 spectra (50 spectra/bundle × 3 bundles) in ATR study were the characterization of fibers longitudinally, and results were given by examining three or four bundle fiber sub-samples in a sample. To verify the M_IR pattern in Figure 4, ATR FT-IR spectra for each of GM-39 and DM-9 fibers were collected on 50 sub-samples and, as illustrated in Figure 6, it showed a Gaussian distribution in M_IR for each sample. Hence, M_IR frequency in Figure 4 could reflect a general fiber maturity occurrence in seed cottons, thanks to significant large sub-sampling of locules in a maturity variable seed cotton sample. Notably, more and random sub-samples in seed cottons might alter the M_IR frequency in Figure 4, as addressed by Kim at al.⁸ that the sample size might cause potential problems in detecting maturity variation within a sample.

Applying a similar concept of separating fibers into low-, mid-, and high-maturity based on respective M_IR value of less than 0.58, 0.58 to 0.80, and greater than 0.80, each sub-sample representing one specific locule fiber was classified into one of 3-M_IR (low-, mid-, and high-) classification classes, and the results are summarized in Table 2. Table 2 indicates that DG3385 cultivar from NM had the highest low-M_IR fiber percentage of 30%, whereas FM1830 from the same location showed the largest high-M_IR number with 88%. In general, two cultivars (DG3385 and PHY444 from NM) showed more than 15% fibers belonging to low-M_IR class, four cultivars (ST5471 from MS, DP1646 dryland from MO, as well as DG3385 and PHY444 from NM) had more than 40% mid-M_IR fibers, and five cultivars (DP1646 and ST4848 from MS, PHY350 and PHY430 from MO, and FM1830 from NM) showed 70% and more high-M_IR fibers.

Table 2.

Comparison of 3-M_IR classification (%) on 11 seed cotton samples

Location/cultivar	3-M_IR classification
Location/cultivar	Low-M_IR (<0.58)	Mid-M_IR (0.58 to 0.80)	High-M_IR (>0.80)
MS
DP1646	2	24	74
ST4848	8	14	78
ST5471	14	50	36
MO
DP1646	6	34	60
DP1646 dryland	2	48	50
NG3522	6	38	56
PHY350	2	24	74
PHY430	2	28	70
NM
DG3385	30	44	26
FM1830	0	12	88
PHY444	18	44	38

In line with expectations, a high-M_IR cotton cultivar identified from the average of 50 locules should have a relatively higher high-M_IR fiber percentage. Indeed, seven high-M_IR cultivars implied a greater high-M_IR fiber percentage than four mid-M_IR cultivars, with a respective high-M_IR fiber percentage mean of 71.4% against 37.5%, mid-M_IR fiber percentage mean of 24.9% against 46.5% ae well as a low-M_IR fiber percentage mean of 3.7% against 16.0%. Cultivars sharing an identical M_IR average might have distinctions in low-, mid-, or high-M_IR fiber percentage. For example, the pair of ST4848 vs. PHY350 cultivar (M_IR average = 0.864 to 0.866) showed a 10-point difference in mid-M_IR percentage (14% vs. 24%), and the pair of DP1646 dryland vs. NG3522 (M_IR value = 0.800 to 0.801) indicated a difference in mid-M_IR count (48% vs. 38%).

Comparison of the DP1646 cultivars grown in MS and MO suggests that they had less than10% low-M_IR fibers in common regardless of the fibers’ growing locations (MS vs. MO) or growing condition (regular vs. dryland). Clearly, DP1646 dryland from MO indicated more mid-M_IR fibers than the other two fiber sets harvested in MS and MO (48% vs. 24–34%), while the irrigated DP1646 from MS produced more high-M_IR fibers than the same cultivar collected in MO (74% vs. 50–60%). In comparison, two PHY cultivars (PHY350 vs. PHY430) from MO were similar to each other in the aspects of low-, mid-, and high-M_IR fiber percentages.

Relating low-, mid-, and high-M_IR fiber percentage from Table 2 against mean M_IR on 11 seed cotton samples in Figure 7 reveals interesting and anticipated patterns. High-M_IR cotton cultivars had a higher high-M_IR fiber percentage, lower mid-M_IR fiber percentage, and lower low-M_IR fiber percentage than mid-M_IR cotton cultivars. With mean M_IR increasing, the high-M_IR fiber percentage rose constantly, while mid-M_IR fiber percentage varied little prior to dramatic reduction near the point of M_IR = 0.80, but low-M_IR fiber percentage reduced gradually. In a recent report, Kim et al.⁴³ proposed a new maturity threshold (0.50) for estimating the immature fiber content (IFC_0.5) from the IA and Cottonscope measurements and observed that the IFC_0.5 values decreased with mean maturity increasing by an exponential function on both developing G. hirsutum TM-1 fibers and the 104 reference materials. Although there were differences terms of immature fibers (i.e., low-M_IR or less mature in this study vs. immature fiber in Kim et al.⁴³), maturity threshold for low-maturity fiber classification (i.e., <0.58 in this study vs. <0.50 in Kim et al.⁴³), and fiber analysis methods (i.e., at bundle fiber level in this study vs. at single fiber level in Kim et al.⁴³), the trend of relating low-maturity fiber count to mean maturity in Figure 7 was similar to those reported for two fiber sets that included the developing TM-1 fibers and the 104 reference materials.⁴³ Like the tendency in Figure 7 and the one reported by Kim et al.,⁴³ AFIS IFC decreased with mean M_AFIS value for 11 ginned cottons (Figure 8).

Comparison of CI_IR frequency on 11 seed cotton samples

CI_IR index was also estimated along with M_IR from the same spectra simultaneously. Identical to the concept of 3-M_IR (high-, mid-, and low-maturity) fiber classification, CI_IR values of less than 58%, 58 to 80%, and greater than 80% were subjectively considered as low-, mid-, and high-crystallinity fiber, respectively. As a whole, a total of two samples (ST5471 from MS and DG3385 from NM) were rated as mid-CI_IR fibers, and the remaining nine cultivars were classified as high-CI_IR fibers.

In general, the data in Table 3 matched that in Table 2. For instance, DG3385 cultivar, with the highest low-M_IR fiber percentage of 30%, had the largest low-CI_IR fiber amount of 18%, and also the FM1830 cultivar, with the largest high-M_IR number of 88%, had the highest high-CI_IR fibers with 90%. One cultivar (DG3385 from NM) showed more than 10% fibers belonging to low-CI_IR class, three cultivars (ST5471 from MS as well as DG3385 and PHY444 from NM) had more than 40% mid-CI_IR fibers, and seven cultivars (DP1646 and ST4848 from MS, DP1646, DP1646 dryland, NG3522, and PHY350 from MO, and FM1830 from NM) showed 70% or more high-CI_IR fibers.

Table 3.

Comparison of 3-CI_IR classification (%) on 11 seed cotton samples

Location/cultivar	CI_IR (%) (average)	3-CI_IR classification (%)
Location/cultivar	CI_IR (%) (average)	Low-CI_IR (<58%)	Mid-CI_IR (58% to 80%)	High-CI_IR (>80%)
MS
DP1646	90.6	0	26	74
ST4848	90.4	0	22	78
ST5471	77.8	8	54	38
MO
DP1646	88.7	0	24	76
DP1646 dryland	88.6	2	26	72
NG3522	85.6	0	30	70
PHY350	91.2	0	20	80
PHY430	85.4	0	32	68
NM
DG3385	73.2	18	52	30
FM1830	96.2	0	10	90
PHY444	80.8	8	42	50

Unlike the fact that three DP1646 cultivars had from MS and MO changed in mid- and high-M_IR fiber numbers (Table 2), they had similar low-CI_IR fiber percentage (<5%), mid-CI_IR fiber amount (24 to 26%), and high-CI_IR fiber percentage (72 to 76%), regardless of fiber grown locations (MS vs. MO) or grown condition (regular vs. dryland). Interestingly, PHY350 cultivar showed lower mid-CI_IR fibers and more high-CI_IR fibers than PHY430 cultivar in two respective readings of 20% vs. 32% and 80% vs. 68%, differing from their similarities in low-, mid-, and high-M_IR fiber percentages between the two cultivars (Table 2). Results underscore that genotype and environment, as well as their interactions, greatly influence fiber maturity and crystallinity properties.

Conclusion

Direct fiber maturity measurement in commercial seed cottons is often difficult because they contain visible trash and cotton seeds that are required to be removed prior to routine fiber maturity testing. With the use of an ATR micro-sampling device that it is capable of analyzing as little as 0.5 mg fiber directly and rapidly without the need either to remove any visible trash materials and cotton seeds or to prepare the fiber samples, this study presented a FT-IR spectroscopy protocol for determining seed cotton fiber maturity and crystallinity at bundle fiber level as well as a method for classing the individual bundle fiber into one of 3-M_IR (low-, mid-, and high-) classifications or one of 3-CI_IR (low-, mid-, and high-) classifications. While the result represented 50 sub-samples in a total sample, it reflected the analyses of 50 different bundle fibers in naturally variable seed cottons. Comparing the fiber M_IR averages on seed cottons to M_AFIS on ginned cottons, it showed a general tendency of M_AFIS increasing along with increasing M_IR on 11 Upland cotton samples with a narrow maturity range. Clearly, the FT-IR result revealed the variation in M_IR frequency, and also provided 3-M_IR and 3-CI_IR classifications among 11 seed cotton samples. Comparison of classification data showed different patterns between M_IR and CI_IR indices within either three DP1646 samples (from MS and MO) or two PHY cultivars (from MO). Results indicated the impact of genotype, environment, and their interactions on fiber maturity and crystallinity development, which in turn, showed that it is essential to apply comprehensive fiber physical and chemical data to understand fiber quality in cotton cultivars.

Footnotes

Disclaimer

Mention of a product or specific equipment does not constitute a guarantee or warranty by the U.S. Department of Agriculture and does not imply its approval to the exclusion of other products that may also be suitable.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Yongliang Liu

References

Robertson

Roberts

BA.

Integrated crop management for cotton production in the 21st century. In: Wakelyn

Chaudhry

(eds) Cotton: Technology for the 21st century. International Cotton Advisory Committee, Washington DC, USA, 2010, pp. 63–97.

Brecke

Banks

Cothren

JT.

Harvest-aid treatments: products and application timing. In: Supak

Snipes

(eds) Cotton harvest management: Use and influence of harvest Aids. The Cotton Foundation, Memphis, TN, USA, 2001, pp. 119–142.

Ayele

Kelly

Hequet

EF.

Evaluating within-plant variability of cotton fiber length and maturity. Agron J 2018; 110: 47–55.

Turner

Sari-Sarraf

Hequet

, et al. Variation in maturity observed along individual cotton fibers using confocal microscopy and image analysis. Text Res J 2015; 85: 867–883.

Anthony

Meredith

Williford

JR.

Neps in ginned lint: the effects of cultivars, harvesting, and ginning practices. Text Res J 1988; 58: 633–640.

Gordon

Cotton fiber quality. In: Gordon

Hsieh

Y-L

(eds) Cotton: Science and Technology. Woodhead Publishing Limited, Cambridge, UK, 2007, pp. 68–100.

Kelly

Abidi

Ethridge

, et al. Fiber to fabric. In: Fang DD, Percy RG (eds) Cotton, 2nd ed. American Society of Agronomy, Crop Science Society of America, and Soil Science Society of America, Madison, WI, USA, 2015, pp. 665–744.

Kim

Delhom

Rodgers

, et al. Effect of fiber maturity on bundle and single-fiber strength of upland cotton. Crop Sci 2019; 59: 115–124.

Long

Delhom

Bange

MP.

Effects of cotton genotype, defoliation timing and season on fiber cross-sectional properties and yarn performance. Text Res J 2021; 91: 1943–1956.

10.

Hequet

Wyatt

Abidi

, et al. Creation of a set of reference material for cotton fiber maturity measurements. Text Res J 2006; 76: 576–586.

11.

Thibodeaux

Evans

JP.

Cotton fiber maturity by image analysis. Text Res J 1986; 56: 130–139.

12.

Pourdeyhimi

Evaluating maturity of cotton fibers using image analysis: definition and algorithm. Text Res J 1994; 64: 330–335.

13.

Bradow

Hinojosa

Wartelle

, et al. Applications of AFIS fineness and maturity module and x-ray fluorescence spectroscopy in fiber maturity evaluation. Text Res J 1996; 66: 545–554.

14.

Paudel

Hequet

Abidi

Evaluation of cotton fiber maturity measurements. Ind Crop Prod 2013; 45: 435–441.

15.

Kim

Delhom

Fang

, et al. Application of the cottonscope for determining fiber maturity and fineness of an upland cotton MAGIC population. Crop Sci 2020; 60: 2266–2279.

16.

Rodgers

Delhom

Fortier

, et al. Rapid measurement of cotton fiber maturity and fineness by image analysis microscopy using the Cottonscope. Text Res J 2012; 82: 259–271.

17.

Rodgers

Naylor

GRS

Cui

, et al. Cottonscope fiber maturity, fineness, and ribbon width measurements with different sample sizes. Text Res J 2015; 85: 897–911.

18.

Montalvo

JG Jr

Von Hoven

Davidonis

Biased experimental fineness and maturity results. Text Res J 2007; 77: 743–755.

19.

ASTM D1442-06. Standard Test Method for Maturity of Cotton Fibers (Sodium Hydroxide Swelling and Polarized Light Procedures). ASTM Intl. West Conshohocken, PA, USA, 2019.

20.

Long

Bange

Gordon

, et al. Measuring the maturity of developing cotton fibers using an automated polarized light microscopy technique. Text Res J 2010; 80: 463–471.

21.

Abidi

Hequet

Cabrales

Applications of Fourier transform infrared spectroscopy to study cotton fibers. In: Nikolic G (eds) Fourier transforms – New analytical approaches and FTIR strategies. InTech, Rijeka, Croatia, 2011, pp. 89–114.

22.

Rodgers

Montalvo

Gayle

, et al. Near infrared measurement of cotton fiber micronaire, maturity and finesses – a comparative investigation. Text Res J 2010; 80: 780–793.

23.

Zumba

Rodgers

Indest

Fiber micronaire, fineness, and maturity predictions using NIR spectroscopy instruments on seed cotton and cotton fiber, in and outside the laboratory. J Cotton Sci 2017; 21: 247–258.

24.

Boylston

Thibodeaux

Evans

JP.

Applying microscopy to the development of a reference method for cotton fiber maturity. Text Res J 1993; 63: 80–87.

25.

Guo

Ouyang

Assessing cotton maturity using distributional parameters of fiber cross-section measurements. Text Res J 2014; 84: 1666–1676.

26.

Matic-Leigh

Cauthen

DA.

Determining cotton fiber maturity by image analysis, part I: direct measurement of cotton fiber characteristics. Text Res J 1994; 64: 534–544.

27.

Thibodeaux

Rajasekaran

Development of new reference standards for cotton fiber maturity. J Cotton Sci 1999; 3: 188–193.

28.

Huang

Image analysis for cotton fibers, Part II: cross-sectional measurements. Text Res J 2004; 74: 409–416.

29.

Liu

Thibodeaux

Gamble

Development of FTIR spectroscopy in direct, non-destructive, and rapid determination of cotton Fiber maturity. Text Res J 2011; 81: 1559–1567.

30.

Rodgers

Delhom

Hinchliffe

, et al. A rapid measurement for cotton breeders of maturity and fineness from developing and mature fibers. Text Res J 2013; 83: 1439–1451.

31.

Santiago

Von Hoven

Hinchliffe

, et al. Examination of cotton maturity and maturity distribution using an infrared focal plane array imaging system. AATCC J Res 2021; 8(1): 14–24.

32.

Kim

Rodgers

Delhom

, et al. Comparisons of methods measuring fiber maturity and fineness of Upland cotton fibers containing different degrees of fiber cell wall development. Text Res J 2014; 84: 1622–1633.

33.

Liu

Kim

H-J

Delhom

, et al. Investigation of fiber maturity measurement by cross-sectional image analysis and Fourier transform infrared spectroscopy on developing and developed upland cottons. Cellulose 2019; 26: 5865–5875.

34.

Liu

Thibodeaux

Gamble

, et al. Comparative investigation of Fourier transform infrared (FT-IR) spectroscopy and X-ray diffraction (XRD) in the determination of cotton fiber crystallinity. Appl Spectrosc 2012; 66: 983–986.

35.

Liu

Kim

HJ.

Use of attenuated total reflection Fourier transform infrared (ATR FT-IR) spectroscopy in direct, non-destructive, and rapid assessment of developmental cotton fibers grown in planta and in culture. Appl Spectrosc 2015; 69: 1001–1010.

36.

Averett

Griffiths

Nishikida

Effective path length in attenuated total reflection spectroscopy. Anal Chem 2008; 80: 3045–3049.

37.

Larkin

Infrared and Raman Spectroscopy: Principles and Spectral Interpretation. Elsevier, Waltham, MA USA, 2011, pp. 27–54.

38.

Široky

Blackburn

Bechtold

, et al. Attenuated total reflectance Fourier-transform infrared spectroscopy analysis of crystallinity changes in lyocell following continuous treatment with sodium hydroxide. Cellulose 2010; 17: 103–115.

39.

Van der Sluijs

MHJ.

Cotton appearance. In: Gordon

Abidi

(eds) Cotton fibers: Characteristics, uses and performance. Nova Science Publishers, Inc. New York, USA, 2017, pp. 135–158.

40.

Abidi

Manike

X-ray diffraction and FTIR investigations of cellulose deposition during cotton fiber development. Text Res J 2018; 88: 719–730.

41.

Abidi

Cabrales

Haigler

CH.

Changes in the cell wall and cellulose content of developing cotton fibers investigated by FTIR spectroscopy. Carbohyd Polym 2014; 100: 9–16.

42.

Liyanage

Abidi

Molecular weight and organization of cellulose at different stages of cotton fiber development. Text Res J 2019; 89: 726–738.

43.

Kim

Delhom

Liu

, et al. Characterizations of a distributional parameter that evaluates contents of immature fibers within and among cotton samples. Cellulose 2021; 28: 9023–9038.

Investigation of fiber maturity and crystallinity information in Upland seed cottons by Fourier transform infrared spectroscopy

Abstract

Keywords

Experimental details

Seed cottons

ATR FT-IR collection, MIR and CIIR determination, and data analysis of seed cottons

Seed cotton ginning, fiber MAFIS and MIR measurement in ginned lint

Results and discussion

Comparison of MIR on 11 seed cotton samples, and MAFIS in ginned samples

Comparison of MIR frequency on 11 seed cotton samples

Comparison of CIIR frequency on 11 seed cotton samples

Conclusion

Footnotes

Disclaimer

Declaration of conflicting interests

Funding

ORCID iD

References

ATR FT-IR collection, M_IR and CI_IR determination, and data analysis of seed cottons

Seed cotton ginning, fiber M_AFIS and M_IR measurement in ginned lint

Comparison of M_IR on 11 seed cotton samples, and M_AFIS in ginned samples

Comparison of M_IR frequency on 11 seed cotton samples

Comparison of CI_IR frequency on 11 seed cotton samples