Synthesis of Azo Disperse Dyes with High Absorption for Efficient Polyethylene Terephthalate Dyeing Performances in Supercritical Carbon Dioxide

Supercritical carbon dioxide dyeing (SCDD) not only enables strong dyeing performance for a versatile range of polymer material but is also regarded as a green chemical media due to its low environmental impact as well as low risk of product denaturation. Over the decades, azo disperse dyes have been revealed to be efficient dyes and represent the wide majority of dyeing material. Azo dyes possess a wide variety of functional groups to optimize dye synthesis and tune the light absorption properties. Using SCDD, end-chain of different lengths, and functional group exhibiting various electronic affinity, six disperse red azo dyes were synthesized to investigate dyeing performances as woven fabric type, color strain, and color fastness after dyeing are discussed. Dye structure synthesized through a coupling reaction was confirmed by 1H NMR and mass spectroscopy. We found that the light absorption wavelength and absorption coefficient value variation are associated to the nature of the functional group. From the color strength values of the polyethylene terephthalate woven after dyeing, we find that the fiber host and dye dopant chemical structure greatly influence the dyeing process by providing enhanced woven, color strain, and color fastness. In comparison with commercial products, our approach not only improves the dyeing process but also guarantees a strong resistance of the dyed product against water, detergent, perspiration, abrasion, and friction.


Introduction
Conventional water dyeing leads to high intrinsic environmental problems such as waste, pollution, high energy, and water consumption. In order to overcome aforementioned drawback, carbon dioxide in supercritical conditions is facile use and exhibit high environmental protection. Supercritical carbon dioxide dyeing (SCDD) is the mainstream of dyeing due to its low environmental impact, non-toxicity, low surface viscosity, and energy, as well as low risk of product denaturation [1][2][3][4][5][6][7]. The SCDD process not only prevents polymeric swelling but also ensures effective diffusion of dye within polymer network while promoting dyestuff recovery after carbon dioxide (CO 2 ) withdrawal. Through the optimization of the free energy, dissolved dye molecules form a single uniform phase with CO 2 . However, crucially, control of the temperature and pressure of the chamber to avoid low solubility and solid-phase separation between host and dye is required. As such, CO 2 A specified quantity of aniline-based diazo precursor and 10mL of ultrapure water are poured into a beaker, then 6 mL (37%) of hydrochloric acid are added under steady stirring and heating. Upon dissolution, a small amount of ice cubes is added to cool down the solution up to 0-5 • C. Then, 0.76 g (0.011 mole) of sodium nitrite aqueous solution is added under continuous stirring. Finally, 4-diethyl Aminobenzaldehyde (IP Solution) is used to check the presence of yellow color, and the reaction undergoes stirring until the disappearance of the yellow color, indicating complete diazotization reaction. Finally, potassium iodide (KI) test paper is used to check the presence of nitrous acid in the reaction solution. If so, traces of sulfamic acid are eliminated until the potassium iodide test paper does not change color and the solution becomes light yellow and clear, indicating a successful diazonium solution. Others diazonium solutions are prepared by replacing 4-methyl aniline with other primary aromatic species.

Dye Fabrication and Purification
A specified quantity of coupling agent is added to the mixture of 7.5 mL cellulose acetate and 2.5 mL ultrapure water at 0-5 • C. Then, carefully drop diazonium solution into the coupling agent solution at 0-5 • C under continuous stirring until reaction's completion. Further, an appropriate amount of sodium acetate is added to the solution to adjust the pH toward weak acidity. Following dye precipitation and filtration, the dye pH is adjusted to a neutral value with ultrapure water. After filtering, TLC is used to determine the presence of impurities in the dye. If so, impurities are washed away with ethanol followed by filtration. This step is repeated until disappearance of impurities. The formula of the developing agent was toluene: ethyl acetate: cellulose acetate = 8:2:1. For each dye, the process follows the same principle at the exception of the diazo components and coupling reagents that are selected to match the targeted dye, as shown in Table 1. Dye synthesis parameters are shown in Table S1 (see Supplementary Materials). the presence of impurities in the dye. If so, impurities are washed away with ethanol followed by filtration. This step is repeated until disappearance of impurities. The formula of the developing agent was toluene: ethyl acetate: cellulose acetate = 8:2:1. For each dye, the process follows the same principle at the exception of the diazo components and coupling reagents that are selected to match the targeted dye, as shown in Table 1. Dye synthesis parameters are shown in Table S1 (see Supplementary Materials).

Characterization
The purified dyes were identified by FTIR, MS, and 1 H NMR. Tetramethylsilane (TMS) was used as a reference of 1 H NMR measurement (Bruker Advance III HD-600 MHz liquid, Bruker Daltonik GmbH, Bremen, Germany). 10mg of the dye sample was added into a flask with 1mL of CDCl3 to dissolve it, followed by injection into 1

Characterization
The purified dyes were identified by FTIR, MS, and 1 H NMR. Tetramethylsilane (TMS) was used as a reference of 1 H NMR measurement (Bruker Advance III HD-600 MHz liquid, Bruker Daltonik GmbH, Bremen, Germany). 10mg of the dye sample was added into a flask with 1mL of CDCl3 to dissolve it, followed by injection into 1 H NMR tube. . For mass spectrometer analysis of dye chemical structure, samples are first ionized through various methods carried out according to their properties. Then, 3mg to 5mg dye sample was placed in the chamber under a high magnetic field to generate a rotary motion, followed by molecular weight structure determination.

Characterization
The purified dyes were identified by FTIR, MS, and 1 H NMR. Tetramethylsilane (TMS) was used as a reference of 1 H NMR measurement (Bruker Advance III HD-600 MHz liquid, Bruker Daltonik GmbH, Bremen, Germany). 10mg of the dye sample was added into a flask with 1mL of CDCl 3 to dissolve it, followed by injection into 1 H NMR tube. For mass spectrometer analysis of dye chemical structure, samples are first ionized through various methods carried out according to their properties. Then, 3mg to 5mg dye sample was placed in the chamber under a high magnetic field to generate a rotary motion, followed by molecular weight structure determination. The MS of dye 019-A of molecular structure C 25

Dyeing
The Figure S1 (see Supplementary Materials) shows the supercritical carbon experimental apparatus was used in this study. The polyester fabric of 75 D/72 F dimension was weighted at 20 g and mixed with a dye solution of 1% concentration and placed into the machine chamber. The dyeing pressure and temperature are set as follows: 3625 psi and 120 • C, respectively, for a dyeing time of 60 min.

Coloring Performance Characterization
The K/S colorfastness is calculated from Kubelka and Munk formula [33] and given as follows in Equation (1): with K being the absorption and S the back-scattering coefficient, R ∞ being the incident light infinite thick layer re-emission fraction. Relative coloring rate is determined as the dye cannot be exhausted during supercritical dyeing. By replacing the dyed material (continued dyeing), the dye can be exhausted. The previously obtained K/S value is divided by the first dyeing each time and accumulation of K/S value is calculated for continued dyeing. The term ( K S ) St is the first K/S value after dyeing and ∑ ∞ n=1 ( K S ) n is the K/S value accumulation after each successive dyeing, with s and t being, respectively and given by Equation (2):

Dyeing Fastness
In accordance with the: ISO 105 C06: 2010 procedure, six kinds of fibers fabric, including cellulose acetate, cotton, nylon, polyester, acrylic, and wool, were sewn to a 10 cm × 4 cm dyed cloths fabric and for all fastness tests as described below. Products were lauded into a stainless-steel bottle (volume: 550 mL, height: 125 mm, diameter: 75 mm) and placed into a washing test machine with ultrapure water, AATCC 1993 Standard Detergent (WOB), as well as 50 steel balls with a constant rotating speed of 40 ± 2 rpm. Following the washing, the sample was washed for 1 min at 40 • C with 100 mL of ultrapure water. Then, fabrics were dissembled and dried at a temperature not exceeding 60 • C.

Water Fastness
In accordance with the ISO 105-E01 procedure, an attached cloth of 10 cm × 4 cm was sewn with a sample cloth and dipped in ultra-pure water at room temperature for 30 min while guaranteeing uniform immersion. Then, the sample was pressed in-between two glasses to remove excess water. The sample was placed between an acrylic resin plate into a preheating test device with a specified load of 5 kg applied onto the test device. The test was conducted at 37 • C for 4 h. Following the end of the test, the load was removed, and the composite sample was extracted and dried at 60 • C.

Perspiration Fastness
All tests were conducted following the procedure explained above. The water was replaced with alkali or acid to mimic sweat, following the ISO 105-E04 specifications.

Color Fastness Abrasion and Rubbing
For dry friction, an AATCC standard white cotton cloth was clamped onto the cylindrical friction head. Then, the sample cloth was tested on the sandpaper of the machine. The downward pressure of the friction head is 9N, and it scrubbed back and forth 10 times at a speed of 1 time/second. For wet friction, a AATCC standard white cotton cloth was soaked in water at room temperature. Then, absorbent paper was used to stabilize the moisture content of the standard white cotton cloth at 65% (±5%). We clipped the standard white cotton cloth to the cylindrical friction head and placed the sample cloth to be tested on the sandpaper of the machine. The downward pressure of the friction head is 9N. During the test, it rubs back and forth 10 times at a speed of 1 time/sec. Finally, the cotton cloths of the two test methods were dried in air at a temperature not exceeding 60 • C. For all fastness tests, the result is recorded after judging the cotton cloth with gray scale or color difference.

Dyes Synthesis and Characterization
In short, disperse azo dyes were obtained from a coupling process with a diazonium salt and coupling component, as shown in Figure 1 and Table 1. The diazonium salt prepared from 5-Methoxy-2-methyl-4-nitroaniline, 4-Nitroaniline, and 2-Methoxy-4-nitroaniline were synthesized through a diazotization reaction. Briefly, aromatic primary amines were used in the production of diazonium salts. The reaction temperature of diazotization was carried out at 0-5 • C to increase diazonium salt stability. At low pH, the sodium nitrite solution is more likely to dehydrate to form nitroso cations, while primary aromatic amines containing hydroxyl or sulfonic acid groups assist the diazotization reaction. The choice of the acid is also primordial to avoid formation of amino cations (NH 3 + ), which reduces the electron density and makes it difficult to carry out nucleophilic substitution reactions. A proper amount of sodium nitrite is required to avoid formation of triazene with primary aromatic amines or quick reaction with the coupling agent producing by-products, thus reducing the diazonium salt purity.
Following the synthesis of diazonium salt precursor, six disperse azo dyes were synthesized using a coupling reaction chemical route. The dye 019 series synthesized from 3-(N,N-diacetoxyethyl)amino-4-methoxyacetaniline and dye 160 series synthesized from 3-(N,N-diethyl)aminopropionaniline coupling component with previously synthesized diazonium component. Traditionally, weakly basic amines are diazotized in acidic media, and the coupling process is required to be conducted in acidic environment for dye stability [23,26]. The coupling reaction was conducted at a temperature maintained at 0 to 5 • C to prevent instability of the diazonium salt. Under acidic conditions, -NH 2 exhibits substantial electron-donating property, favorable to the coupling reaction. Besides working at low temperature and acidic media, diazonium salt stability is also influenced by the strength of the salt substituent electron-withdrawing group, where a stronger one will accelerate the coupling rate at the expense of the salt stability. Our dyes are composed of azo groups as the primary chromophore, combined with multiple auxochromic groups to endow the dye with a particular color and dyeing capability. Interestingly, the presence of the methoxy group at the meta-position originates from the nature of coupling agent substituents and pH of the media. Electron withdrawing groups weaken the electron density, whereas electron repulsive groups enhance the electron density toward the coupling agent and favor reactions at the meta-position of the substituent. More details are given about the synthesis principle and objectives are given in the Supplementary Materials. As shown in Figure 1, three sub-series of dye 019 and 160 series were synthesized by tuning the position of the electron withdrawing and repulsive group to endow disperse dye with specific properties, including high light absorption and strong color fastness. agent and favor reactions at the meta-position of the substituent. More details are given about the synthesis principle and objectives are given in the Supplementary Materials. As shown in Figure 1, three sub-series of dye 019 and 160 series were synthesized by tuning the position of the electron withdrawing and repulsive group to endow disperse dye with specific properties, including high light absorption and strong color fastness.  Further, we conducted FTIR, as shown in Figure 2, for clarity, all values are given for the following dye order 019-A, 019-B, and 019-C. We find that for the dye series 019, all dyes present a signal at 3382. 32

Optical Characterization of Dyes
Synthesized on a D-π-A model with electron-donating (D) or electron-accepting (A) functional groups, disperse dyes exhibit various color deepness, strength, and shades in correlation with the strength of the electron withdrawing group. The disperse dyes were dispersed with ethanol at a concentration of 10 −5 g/mL for UV-Vis experiment with the In conclusion, our dyes display highly satisfying chemical properties while having high purity and yield. Therefore, with dye structure integrity being demonstrated, the presence of functional group at the desired position is expected to significantly influence the optical properties of disperse dyes.

Optical Characterization of Dyes
Synthesized on a D-π-A model with electron-donating (D) or electron-accepting (A) functional groups, disperse dyes exhibit various color deepness, strength, and shades in correlation with the strength of the electron withdrawing group. The disperse dyes were dispersed with ethanol at a concentration of 10 −5 g/mL for UV-Vis experiment with the maximum absorption wavelength (λ max) and absorption coefficient (ε) are shown in the Table 2. As shown, in Figure 3, the maximum absorption wavelength is ranging around the same maximum region around 513-562 nm thus emitting red light, differences are observed due to the introduction of end-group grafted onto aromatics group and side-chains of different length. The maximum absorption wavelengths of dyes 019-A and 160-A are 513 nm, 512 nm, respectively, the maximum absorption wavelengths of dye 019-B and 160-B are 519 nm and 516 nm, and the maximum absorption wavelengths of dye 019-C and 160-C are 562 nm and 516 nm, respectively. The low light absorption wavelength shift means that the length of the side-chains does not significantly influence on the maximum absorption wavelength. Besides, for each dye series, ε value are over 35,000 L·mol −1 ·cm −1 up to 216,000 L·mol −1 ·cm −1 . It is admitted that a higher ε value is associated with a lower number of dyes required to dye fabric as well as a much higher electron repulsion effect [27]. Zhang et al.'s approach led to an ε to be as high as 31,000 L·mol −1 ·cm −1 . While the hydroxyl group endows dye with water compatibility, it is expected that the hydroxyl group being a weak electron withdrawing group reduces dyeing efficiency [26]. By replacing hydroxyl with stronger electron withdrawing group such as ester and urethane group, the hypsochromic shift increase, translating a stronger electron pulling effect [27]. Furthermore, significant variations of ε are observed within and between dyes series, with the dyes 160 series exhibiting the highest ε value overall. First, the methyl group at the aromatic ring para position acts as an electron, donating substituent pushing electrons toward the chromophore. Such a repulsive effect causes the bandgap to widen in energy, translating a shift to the shorter wavelength of 4-6 nm. Second, high ε is attributed to the electron withdrawing methoxy group at the meta of the aromatic ring so that the electron pulling effect from the amine group is enhanced for dyes series A. This is confirmed, as dye 019-C and dye 160-C maximum wavelength of absorption has drastically shifted toward the red region, as they do not possess methoxy group at the meta-position of the phenyl ring nor methyl group. According to Cinar et al. [36], azo disperse dye constructed on -C-N=N-C structure with phenyl rings has been observed to have allowed or forbidden transitions due to symmetry consideration. Cinar et al. not only assigned their azo disperse transition at 490 nm to n-π* but also highlighted that the band-gap of the chromophore could be decreased with increased π-electron delocalization. However, Lee et al. highlighted that the presence of a methoxy group onto the coupling component at the meta-position induces a bathochromic and hypochromic effect [37]. Therefore, the electron pushing effect is weakened. Herein, the molecule functional group engineering results in a red-shifting of the absorbance while preserving the n-π* transition, with the n orbital corresponding to the lone pair of electrons of the N atoms. Our observation corroborates this statement as our dye series 019 possesses a methoxy group at the meta-position of the coupling component. Dye 019 series wavelength shifts much more significantly toward red region with lower ε value in comparison to the dye 160 series. This observation highlights the importance of the methoxy group at the meta-position onto the aromatic group with respect to the D-π-A dye model to significantly enhance the light absorption performances. Overall, all dyes are suitable for dyeing. [27]. Zhang et al.'s approach led to an ε to be as high as 31,000 L·mol −1 ·cm −1 . While the hydroxyl group endows dye with water compatibility, it is expected that the hydroxyl group being a weak electron withdrawing group reduces dyeing efficiency [26]. By replacing hydroxyl with stronger electron withdrawing group such as ester and urethane group, the hypsochromic shift increase, translating a stronger electron pulling effect [27]. Furthermore, significant variations of ε are observed within and between dyes series, with the dyes 160 series exhibiting the highest ε value overall. First, the methyl group at the aromatic ring para position acts as an electron, donating substituent pushing electrons toward the chromophore. Such a repulsive effect causes the bandgap to widen in energy, translating a shift to the shorter wavelength of 4-6 nm. Second, high ε is attributed to the electron withdrawing methoxy group at the meta of the aromatic ring so that the electron pulling effect from the amine group is enhanced for dyes series A. This is confirmed, as dye 019-C and dye 160-C maximum wavelength of absorption has drastically shifted toward the red region, as they do not possess methoxy group at the meta-position of the phenyl ring nor methyl group. According to Cinar et al. [36], azo disperse dye constructed on -C-N=N-C structure with phenyl rings has been observed to have allowed or forbidden transitions due to symmetry consideration. Cinar et al. not only assigned their azo disperse transition at 490 nm to n-π* but also highlighted that the bandgap of the chromophore could be decreased with increased π-electron delocalization. However, Lee et al. highlighted that the presence of a methoxy group onto the coupling component at the meta-position induces a bathochromic and hypochromic effect [37]. Therefore, the electron pushing effect is weakened. Herein, the molecule functional group engineering results in a red-shifting of the absorbance while preserving the n-π* transition, with the n orbital corresponding to the lone pair of electrons of the N atoms. Our

Apparent Colorfastness K/S
Conventional water polyester fiber dyeing with disperse dyes requires pre-treatment and being coated with auxiliary agents (such as methyl salicylate, benzoate, chlorobenzene, methyl naphthalene). On the contrary, SCDD with disperse dye does not require pretreatment and can be dyed simply by drying and grinding. Disperse dyes are hardly soluble in water, and are more likely to be dispersed in supercritical CO 2 while avoiding phase separation. Therefore, the crystallinity ratio of the PET fibers directly impacts the dye penetration efficiency into the PET fibers [38,39]. Moreover, CO 2 does not penetrate the PET fabric due to hydrophobicity incompatibility, therefore facilitating dye diffusion within the amorphous PET part. During dyeing, the dye is first absorbed onto the fiber boundary layer during dye flowing, followed by inward diffusion within the fiber boundary layer, triggering intermolecular interaction between fiber host and dye. The diffusion rate is proportional to the temperature and dye concentration. Therefore, increasing the dyeing temperature increases the pores in the amorphous area of the polyester fiber and reduces the resistance of the dye molecules to diffuse into the fiber, thereby increasing the dye uptake rate and reducing the diffusion time. One of the main advantages of PET fibers is its low T g facilitating chain motion and wider amorphous area.
The PET fabric was mixed with dye solution at 1% concentration and placed into the machine chamber, followed by a SCDD process for 60 min at 120 • C. A high-temperature process facilitates the penetration of dyes within non-crystalline area of PET fibers. To assess the dyeing performances, the dye absorption efficiency of dyed PET fabric is calculated using the K/S formula. We find that our dyes K/S apparent color density range from 14.78 (dye 019-C) to 20.02 (dye 160-A). Table 3 shows that dye 019-B has a higher apparent color concentration in the 019 series (16.376), and relative coloring rate is 89.30%, higher than other 019 series dyes. The apparent color concentration of dye 160-A is higher (20.02), and its first relative coloring rate is 88.82% better than other 160 series dyes and 019-C has the highest first relative coloring rate among all dyes. Based on the above description, 019-A, 019-B, 019-C, 160-A are more suitable for SCDD dyeing than other dyes of their corresponding series. We find that the methoxy group, as shown previously, also significantly impacts the apparent color density by improving the electron pulling effect of successfully incorporated dye into PET. Disperse azo dye with strong electron withdrawing groups provide a higher electron pulling effect, leading to improved color fastness [27].
It shows that the dye's molecular structure directly impacts the color rendering yield following the dyeing process. Moreover, it is known that the addition of side chain of significant length such as alkyl chains to the dye can increase the apparent color density, the first relative coloring rate, and influence the colorfastness properties after dyeing [40,41]. Freeman et al. [39] demonstrated that the length of the alkyl chains attached to the phenyl ring do not exhibit a significant influence onto the polylactide fibers dyeing performances but greatly influence the dyeing fastness performances. Considering that the dye 019 series possess longer end-chain compared to the dye 160 series, the difference in color fastness originates from the excessive length of the carbon chains, reducing penetration and diffusion of dye into the amorphous area of the polyester fiber. As a result, both the apparent color density and the first relative coloring rate significantly decrease. Therefore, the dye's solubility in supercritical CO 2 is enhanced through the introduction of an aniline component for the synthesis of diazo-based disperse dye. However, long side chains are likely to mitigate the dye diffusion within a non-crystalline area of the fiber, as increasing the length of the long-chain group makes the dye molecules larger.

Color Fastness Properties
Different textile fibers fabrics have repeatedly been exposed to various environment in order to evaluate the dyes resilience within fibers fabric in accordance with suitable ISO procedure (We refer to each table description for the ISO procedure name). We have used cotton, nylon, polyester, wool, acrylic, and cellulose acetate fabric sewed onto the dyed stuff. According to ISO test procedure regulation, the washing and dyeing fastness scale range from 1 to 5, with 4-5 associated with excellent wash fastness output. In brief, fabrics were sewn together before being exposed to detergent, water, alkali, and acidic environments. As shown in Table S5 (see Supplementary Materials), the color fastness to washing of dyes 160-A, 160-B, and 160-C are grade 2 when attached to the nylon cloth, followed by exposition to a standard detergent. Dyes 160-A and 160-B reach grade 2-3 when sewed with cellulose acetate mat. The washing fastness of other dyes is above grade 3-4, which meets the commercial standard. Water washing and color fastness has also been investigated, and fabric was exposed to ultrapure water before being dried following ISO 105 E01:2010 specification, as shown in Table 4. The fastness is grade 3 or above for any sewed dyed fabric, which also meets the commercial standard. To go further, we have assessed the dyed fabric resistance to perspiration through exposition to acid (Table 5) and alkali (Table S6, see Supplementary Materials) environments. The observed color fastness after dyeing is all above grade 4, which is in line with commercial standards for any of the fibers and environmental exposure. Finally, dyeing fastness to abrasion and rubbing was investigated and all dyes are above 4, which meets commercial standards as shown in Table 6. We find that regardless of the chemical structure that impact the dyes uptakes and performances, our dyes are strongly impregnated into our fabrics and strongly protected against external degradation.   Table 6. Color fastness to abrasion and rubbing after dyeing according to the AATCC 8 protocol.

Conclusions
In this study, we have successfully synthesized six disperse azo dyes through a facile chemical route consisting of using a coupling component with a dye precursor. We confirmed the dyes structure by 1 H NMR and MS. Following synthesis, dyes absorption color was confirmed to be red. Not only did the end-chains length only weakly influence the color absorption, but the ortho-position of the methoxy group onto the aromatic group strengthened the electron repulsion effect, therefore granting our dyes with high molar extinction coefficient. Further, the dyes 019-A, 019-B, 019-C, and 160-A exhibit the best PET fabric dyeing performances using supercritical CO 2 . According to the apparent color concentration of polyester fabric after dyeing, aniline increases the dyes solubility in supercritical CO 2 , while excessive end-chain length negatively impacts dye performances. Detergent, water, alkali, and acid color fastness test were performed and highlight the superior performances of our synthesized dyes. Abrasion and rubbing test similarly highlight the high diffusion of dye within PET fabric and confirm our dyes meet commercial standard.