The Film-Forming Characterization and Structural Analysis of Pectin from Sunflower Heads

A natural low-methoxyl pectin (termed AHP, DM= 25:9%) was extracted from dried heads of sunflower and showed better filmforming performance blended with hydroxypropyl methylcellulose (HPMC). The solutions and films of different HPMC/AHP blends were characterized by viscosity, transparency, mechanical properties, loss on drying, water drop penetration time (WDPT), disintegration and SEM. In order to analyze the structure-property relationship of film forming, AHP was separated by ion-exchange chromatography and characterized. The results showed that the blends were immiscible, but the formation of AHP gel would give the blended film better mechanical properties. AHP was fractionated into one neutral fraction and two acidic fractions (AHPA-1 and AHPA-2). The analytical results showed that AHPA-1 and AHPA-2 were identified to be homogalacturonan(HG-) rich pectins with low DM, and the molecular weights of them were estimated to be 106 kDa and 226 kDa, respectively. Due to the high content of the HG domain, low DM and high molecular weights, AHP had excellent gelling properties induced by Ca and was added to improve the film-forming properties of HPMC and to develop plant hollow capsules.


Introduction
The medicinal capsules provide a better solid dosage form for drugs with low compressibility, slow dissolution and bitter taste, and they require disintegration in the stomach as quickly as possible in most cases. Gelatin has been used in the production of capsules since the early 19th century and still dominates the pharmaceutical capsule materials because it exhibits good film-forming and gas barrier properties [1]. However, the defects of gelatin capsules are gradually exposed with the development of science and technology and the progress of society. The higher price of the raw materials and well-recognized shortcomings of gelatin capsules, such as animal-derived ingredients, lower soft temperature and the instability of moisture in gelatin [2,3], have led to attempts to use replacement substances, such as hydroxypro-pyl methylcellulose (HPMC) [4], carrageenan [5] and modified starches [6]. Although there are many patents and publications on the developing various substances for capsules, few non-gelatin capsule products have entered the market. The reasons for the less commercial success include the higher price due to the requirements for new processing facilities, as well as more complex processing conditions [7,8]. Due to the lower film-forming and curing properties, the preparation process of the starch capsule is difficult to control, and the brittleness and hygroscopicity of the product are high [9]. The capsule prepared with HPMC as a material has lower hardness, so it is hard to pull out from the mold without being destroyed [10], while the hardness of HPMC capsules could be improved by adding tamarind polysaccharides, carrageenan, aminated low-methoxyl pectin, gellan and broom alginan [11]. Due to the good gel properties without high sugar content, low-methoxyl pectin added to the HPMC capsule will give the capsule higher hardness and stability. And the HPMC/AHP blended capsule will be more suitable for the special requirements of patients with obesity, diabetes and hyperlipidemia.
Pectin, a kind of acidic polysaccharides, mainly exists in plant cell walls [12] and linings as supporting materials. It has been widely applied as the thicker [13], stabilizer [14], emulsifier [15], suspending agent [16] and gelling agent [17], and received special attentions from food [18], cosmetics and medicine industries [19]. The main molecular structure of pectin was composed of homogalacturonan (HG), rhamnogalacturonan I (RG-I), rhamnogalacturonan II (RG-II) and xylogalacturonan (XGA) [20]. HG is the linear polymer comprised of the 1,4-linked α-D-galacturonic acid (GalA) backbone [21]. RG-I is made up of the repeating disaccharide [-4)-α-D-GalpA-(1-2)-α-L-Rhap- (1-], and its side branches composed mainly of neutral sugars, arabinan, galactan and arabinogalactan with highly variable structure and composition [22]. RG-II has a backbone of HG with complex and conservative side chains attached to the GalA residues [23]. XGA has a backbone of HG with xylopyranose as side chains attached to C-3 of GalA residues [24]. With different proportions of carbomethoxy [25], it could be divided into high-methoxyl pectin (HMP, DM > 50%) and low-methoxyl pectin (LMP, DM < 50%). Compared with HMP, the quick gelling properties induced by Ca 2+ and over a wider range of pH values without sugar of LMP are appropriate for low calorie or sugar-free food additives and ingredients of health-care medicines [19]. Up to now, most of the commercial pectin produced from citrus peel, apple pomace [26], sugar beet pulp [27], orange [28], lemon [29], banana [30] and peach [31] were HMP. And HMP is usually served to manufacture LMP by chemical or enzymatic treatment to hydrolysis of the methyl esters from galacturonic acid C-6 carboxyl groups. However, this kind of synthesized LMP is costly and has hazardous residuals brought in by chemical and enzymatic treatment. Mature sunflower (Helianthus annuus L.) heads contained 15-25% (w/w) natural LMP with degree of methyl esterification ðDMÞ ≤ 26% [32,33]. Sunflower heads were usually discarded as wastes from the sunflower oil processing regardless of the high application value for recycling. Since Colin and Lemoine [34] reported the presence of a kind of special pectin from sunflower heads, several studies have been conducted to obtain a better understanding of its properties, and it was verified that LMP from sunflower heads had superior gelling properties induced by Ca 2+ [35]. Whether LMP from sunflower heads has good film-forming properties as capsule ingredients has not been deeply investigated.
The objective of this research is to develop the HPMC/AHP blends as a hard capsule to overcome the lower hardness of the HPMC capsule. Firstly, the sunflower heads pectin (AHP) was extracted by 0.2% oxalic acid solution at 80°C for 1 h and then precipitated with ethanol. Then, the films with different ratios of HPMC/AHP were prepared. And the viscosity of different HPMC/AHP blended solutions, the mechanical properties, loss on drying, water vapor permeability and water drop penetration time of the films were studied. The morphologies and compatibility of different films were investigated by transparency rations and SEM. Finally, in order to find out the structure-property relationship of this novel pectin, AHP was separated by an ionexchange chromatography and purified by size exclusion chromatography into one neutral fraction and two homogeneous acidic fractions. The two acidic fractions were furthermore characterized by sugar composition analysis, highperformance gel permeation chromatography (HPGPC), FT-IR and NMR spectroscopy, successively. The results are anticipated to be not only stimulated the development of plant hollow capsule industries but also dealt with the recycling and reusing problems of sunflower heads.

Results and Discussion
2.1. Preparation AHP from Sunflower Heads. AHP from sunflower heads was extracted with 0.2% oxalic acid solution at 80°C for 1 h and precipitated with 60% ethanol with the yield up to 18.4%. Monosaccharide composition analysis by highperformance liquid chromatography (HPLC) indicated that AHP consisted of GalA (92.3%), Rha (3.7%), Gal (2.1%) and Ara (1.8%) as shown in Table 1.
2.2. The Viscosity of HPMC/AHP Blends. The viscosity of film-forming dispersions is critical to film-forming properties. NDJ-1 type rotary viscometer was used to determine the viscosity of different HPMC/AHP blended solutions. Results showed that AHP had an important impact on viscosity of the dispersions ( Table 2). The viscosity of HPMC solution was 2130 mPa•s, and the viscosities changed with different content AHP added. When the content of AHP was 5%, 10%, 15% and 20% (AHP/HPMC mass ratio), the viscosity was 2055 mPa•s, 2785 mPa•s, 2675 mPa•s and 2530 mPa•s, respectively. Liquid viscosity between 2500 and 3000 mPa•s was suitable for the preparation of plant hollow capsules [36]. So, 10% AHP, 15% AHP and 20% AHP blended solutions could be used to prepare plant hollow capsules.  Table 3 showed that the LOD of HPMC/AHP films were lower than that of conventional gelatin hard capsules (moisture content between 13% and 16%) [37] and the HPMC film. However, the increasing of addition of AHP had little influence on the values of loss on drying of composite films. It manifested that even 5% AHP could effectively eliminate the disadvantages of high loss on drying of other capsule ingredients.

Mechanical
Properties. The mechanical properties of the films with different HPMC/AHP blends were studied by tensile testing. The tensile strength (TS) and percentage elongation at break (E) are used to describe the mechanical properties of the films. The results in Table 3 indicated that the tensile strength decreased slightly and then increased with the AHP content increasing. For elongation properties, the HPMC film had the highest value which was consistent with 2 International Journal of Polymer Science the measurement results in the previous article 13:8 ± 4:2% [37]. And the elongation of the blended films gradually decreased with the AHP content increasing. The tensile strength of the film added 10% AHP was the lowest, which may be caused by two factors: HPMC/AHP blends were an immiscible system, and more importantly, the concentration of AHP was not high enough to form a stable gel structure. With pectin proportion increasing from 10% to 20%, the films became tougher, because the typical "egg box" gel structures [38] formed by AHP and Ca 2+ became more stable and tight. However, the formation of AHP gel would reduce the flexibility of films and increase the brittleness of films which was characterized by the increase of the tensile strength of the blended films and the gradual decrease of the elongation at break.

The Water Vapor Permeability (WVP).
Water vapor permeability (WVP) is an important index to evaluate the capsule properties of protecting inner drugs from water. The water content is one of the key factors affecting the stability of solid pharmaceutical preparations, which could cause hydrolysis reaction, oxidation reaction and even mildew. The lower the WVP, the better water resistance capsules have. The results for WVP of different blended films are also shown in Table 3. The HPMC film had a lower WVP, and WVP decreased with the AHP content increasing up to 5% then increased with more AHP. And the permeability coefficient of blended films was the same as that of gelatin (0.24 g•mm•m -2 •h -1 •kPa -1 ) [39], which demonstrated that the films had a compact structure and met the demand for WVP as plant capsule ingredients.

Water Drop Penetration Time (WDPT).
WDPT is the limitation time of capsule disintegrating, which was a physical dissolving process of pharmaceutical preparations before absorption. It indirectly reflects the water solubility of HPMC/AHP blended films. The value of WDPT/d is smaller, and the permeability of the films is better. As shown in Table 3, the WDPT/d of blended films decreased with the AHP content increasing. The results showed that AHP was more hydrophilic, and the addition of AHP to HPMC would improve the water solubility and disintegration of blended films, which is significance for capsule materials.
2.3.5. Transparency. The pure HPMC solution is reasonably clear, and the transparency ratios of the blended films measured by UV spectrum are shown in Table 3. It is seen that the transparency ratios of films decreased with the AHP content increasing. This phenomenon indicated that the HPMC/AHP blended system was immiscible since phase separation results in the low transparency [10]. And with the increase of the AHP content, the crosslinking degree between AHP and Ca 2+ increased, which promoted the low-methoxyl pectin to form a typical "egg-box" gel structure. According to Yang [40], films with a higher crosslinking degree tend to exhibit decrease in transparency due to a decrease in polymer interchain spacing, allowing less light to pass through the film. Therefore, it could be speculated that the crosslinked gel structure in blended films increased with the AHP content increasing.
2.3.6. Microstructures. SEM was used to investigate the microstructures of the surfaces and the cross-sections of the blended films, as exhibited in Figure 1. The films of pure HPMC exhibited a smoother and continuous surface, and the surface roughness increased with the AHP content increasing. Some protrusions were observed on the surfaces of the films containing 15% and 20% AHP, and the density of these protrusions increased with the AHP content increasing. The results probably caused by the formation of a more rigid structure-"egg-box" gel structure-which may also be responsible for the brittle aspect of blended films. The phenomenon indicates that HPMC and AHP are two phases, but the individual surfaces generally present as a continuous phase without indication of phase separation, which means HPMC and AHP are compatible although they are immiscible [10]. The transversal sections showed that the addition of AHP affected the morphology of the blended films, since the increase of the AHP content to 10% resulted in the decrease of continuous and packed arrangement. However, the structure of film blended with 20% AHP became denser, which was the reason for its maximum tensile strength. These results were consistent with the mechanical properties of the blended films.

Physicochemical Properties, Fractionation and Structural
Analysis of AHP 2.4.1. Fractionation of AHP. According to the method mentioned above, the DM of AHP was 25.9% (DM<50%), indicating that AHP was a low methyl esterification pectin. Since AHP has satisfying film-forming properties, in order to clarify the structure-function relationships, the fractionation and structure analysis of AHP are necessary to be studied. Before fractionation on a semipreparative scale, AHP  3 International Journal of Polymer Science was analyzed on an analytical DEAE-cellulose column eluted by a linear gradient of NaCl (Figure 2(a)). The profile showed an unbound portion eluted with water and a bound portion eluted with linear gradient of NaCl. The elution profile indicated that AHP contained a slight amount of neutral polysaccharides and a large proportion of acidic polysaccharides, and the peaks of two acidic polysaccharides appeared at 0.35 M and 0.5 M NaCl. So, we separated AHP on a preparative DEAE-cellulose column by elution with a stepwise gradient of NaCl. With these steps, AHP was separated into three fractions (Figure 2(b)): AHPN eluted with water (6.9%, no further study), AHPA-1 (38.4%) and AHPA-2 (34.2%) eluted with 0.35 M and 0.5 M NaCl, respectively.
The molecular weight distributions of the two acidic fractions on sepharose CL-6B column were shown in Figure 3. Both AHPA-1 and AHPA-2 showed single and relatively symmetrical peaks, and the distribution of total sugars was consistent with that of uronic acid. The elution profiles from the TSK column showed single and symmetrical peaks for the two fractions ( Figure 4). These results indicated that the two samples were homogeneous fractions related to molecular weights. The molecular weights were approximately deduced from the calibration curve of dextran standards by HPGPC to be 103 kDa (AHPA-1) and 226 kDa (AHPA-2).
The yield and monosaccharide compositions of the collected fractions from AHP are listed in Table 1. Both AHPA-1 and AHPA-2 mainly contained GalA and a small amount of Rha, Gal, and Ara. This result indicated that the two fractions were composed of pectic domains. The ratio of Rha/GalA determined for AHPA-1 was 0.051, which was among the RG-I range from 0.05 to 1.0 [41]. It suggested that AHPA-1 mainly contained HG domains and trace RG-I domains. AHPA-2 contained high amounts of GalA (96.6%) and low ratio of Rha/GalA (0.035), which suggested that AHPA-2 mainly contained HG domains.

Structural
Analysis of AHP Fractions. The FT-IR spectra of the three samples (AHP, AHPA-1 and AHPA-2) are shown in Figure 5. The major peak centered at 3400 cm -1 is due to stretching of the hydroxyl groups. The peak at 2800 cm -1 is due to the C-H stretching of the CH 2 groups, and the peak at 1412 cm -1 corresponds to vibrations of the O=C-O structure. The three samples show high wavenumber values, which constitutes the "fingerprint" region, specific for each polysaccharide. These peaks are originated from complex interacting vibrational modes and as a consequence, it is quite impossible to assign a certain peak of this region to a certain specific atom group vibration. The peak at 1740 cm -1 corresponds to the absorption of the esterified carboxylic groups of the pectin molecules, while the peak at 1610 cm -1 is attributed to the absorption of the carboxylate anions. The ratio of the area of the peak at 1740 cm -1 over the sum of the areas of the peaks at 1740 and 1610 cm -1 can be proportional to the degree of esterification of pectin. It could be seen from the ratios of the peak area that the three samples had low DM, which was consistent with the results of DM determination. And the fraction AHPA-2 had a lower DM than that of AHPA-1.
The 13 C NMR spectra of AHPA-1 and AHPA-2 were quite similar ( Figure 6) and showed predominant resonances from carbons in free galactopyranosyl uronic acid residues in accordance with those previously observed in HG-rich pectins [42]. The signals for C-1 to C-6 of unesterified GalA residues appeared at 97.95 (C-1), 67.09 (C-2), 67.78 (C-3), 76.87 (C-4), 70.23 (C-5) and 174.33 (C-6) ppm. C-1 to C-5 from esterified GalA residues did not give obvious signals because of the low DM, while the methyl ester carbonyl carbons and methyl group showed the diagnostic signals at 169.94 and 51.87 ppm, respectively. Combining the NMR analysis and monosaccharide composition analysis, it is deduced that AHPA-1 and AHPA-2 are HG-rich pectins with low DM, and AHPA-2 had a lower DM.
In conclusion, AHP was a natural low-methoxyl pectin extracted from sunflower heads. AHP had low DM (25.9%), high GalA content (92.3%), high molecular weights (103 kDa for AHPA-1 and 226 kDa for AHPA-2), and mainly composed of HG domains, hence the conformation of AHP might be more stiff rod-like [39]. AHP had superior gelling properties and could formed the typical "egg-box" gel structure by the crosslinking with Ca 2+ . The properties of the HPMC/AHP blended films were related to the degree of crosslinking in the gel structure and the homogeneity of the film-forming systems. The results of the transparency test (Table 3) and SEM (Figure 1) showed that the homogeneity of blended films decreased with the increasing AHP content, which would lead to the decrease of tensile strength of the films. On the other hand, with the increase of the AHP content, the tensile strength would increase as the crosslinking in the gel structure with Ca 2+ . So, the tensile strength of the blended showed two-stage variation pattern (decreased slightly and then increased) with the increase of the AHP content. That of the film blended with 10% AHP was the lowest, because the HPMC/AHP blends were immiscible, and the concentration of AHP was not high enough to form a

4
International Journal of Polymer Science stable gel. And then, the addition of AHP would increase the brittleness of films and reduce the flexibility of films, which was characterized by the increase of tensile strength and the decrease of percentage elongation at break.

Isolation of Pectin from Sunflower Heads.
Crushed deseeded sunflower heads (500 g) were soaked in 80% ethanol for 3 h to remove some colored materials, oligosaccharides, and other small molecule materials. The pretreated sunflower head was immersed 20 L 0.2% oxalic acid solution and extracted in 80°C for two times (1 hour each time). The aqueous filtrates were combined and concentrated to 2 L. 95% ethanol was added into the filtrates until the ethanol concentration reached 60% of total volume. The precipitated pectin was collected after centrifugation and dried in a vac-uum drying oven. Then, collected pectin was dissolved in water (10% w/v) to remove the insoluble substances by centrifugation. The supernatant was treated with the Sevage reagent (1 : 4 n-butanol/chloroform, v/v) to remove proteins. The water phase was dialyzed against distilled water and lyophilized to obtain pectin named AHP.

Solution Preparation and Viscosity Measurement.
Commercial HPMC and pectin from sunflower head AHP dispersions (100 g) were prepared. The appropriate amount of AHP and HPMC (the sum of AHP and HPMC was kept constant at 20 g, and the mass ratios of AHP/HPMC were fixed on 0 : 100, 5 : 95, 10 : 90, 15 : 85, and 20 : 80) and 2% plasticiser glycerol were dissolved in distilled water. The pH of dispersions was adjusted to 4.8, and 1 mL Ca 2+ solution was added to the above dispersions in accordance with the 30 mg/g (Ca 2+ /AHP mass ratio). The suspension was heated

Films Preparation.
The films were prepared by the traditional pouring method. Each biopolymer dispersion was poured into square glass petri dishes (15 cm × 15 cm = 225 cm 2 ), and the solvent was evaporated at room temperature overnight. The films produced were removed and stored in sealed containers at room temperature and 75% RH for at least one week prior to use. Vernier caliper was used to measure the thickness of the films. The average thickness of the film at 10 random positions was adopted as the mean thickness of film.

Properties of the Films
3.4.1. Loss on Drying. The loss on the drying test of the films was conducted as follows [43]. 0.5 g of the film sample was accurately weighed and placed into a constant weighing bottle and dried in a chamber at 105°C for 12 h. After drying, the bottle was cooled in desiccator containing silica gel as an absorbent, and then the bottle was reweighed accurately. Loss on drying percent was calculated according to Eq. (1).
Loss on drying % ð Þ = undried weight − dried weight ð Þ /undried weight × 100: percentage elongation at break (% E) of the preconditioned films were determined at room temperature using a Texture Analyzer (TA-XT2, Stable Micro System Co. Ltd., Surrey, England) [44]. Films were cut into strips (10:0 cm × 2:5 cm) and mounted between the tensile grips of the instrument. The initial grip spacing and crosshead speed were set at 5.0 cm and 1 mm/s, respectively. Measurements were repeated at least 3 times. A microcomputer was used to record the stress-strain curves. The tensile strength was expressed as the maximum force at break per initial crosssectional area of the film. The % E was calculated by dividing the initial distance between the grips and the elongated distances until the film breaks.

The Water
Vapor Permeability (WVP). The water vapor permeability coefficient through the films were determined gravimetrically at 25°C according to the ASTM method [43]. Film samples were sealed over a circular opening of a permeation cell filled with anhydrous calcium chloride maintaining a stagnant air gap of less than 1 mm below the film. These cells were individually kept in hermetically closed chambers (≈500 mL) containing a saturated solution of sodium chloride at 25°C in order to maintain a RH difference of 75%. The salt solution in the bottom was occasionally agitated through a magnetic stirrer. After the system reached steady-state conditions (≈2 h), the cell weight was measured at every 12 h for 5 days. The WVP was calculated using Eq.
where d is the film average thickness (mm), m is the permeation rate (g/h) calculated by linear regression of the mass gain versus time, A is the permeation area (1.521×10 -3 m 2 ), ΔRH is the difference in relative humidity (0.75), and Pw is the partial water vapor pressure at test temperature (3.167 kPa).

Water Drop Penetration Time (WDPT).
Relative to the disintegration time of the capsule, water drop penetration time (WDPT) was measured in the study [45]. Five drops of distilled water (20°C) were applied to the film surface using a hypodermic syringe. The penetration time of the drops was recorded until the film leakage and the median penetration time were taken as representative of the WDPT of each sample. WDPT/d was described for the WDPT per film thickness, where d is the thickness of composite films (mm).
3.4.5. Transparency. A UV (WFZ UV-3802) spectrum was used to measure the transparency of the different blended films, which were cut into strips (9 mm × 10 mm) and placed in a 10 mm × 10 mm square sample container for measurement. The transparency of films was measured at a wavelength of 206 nm [10]. Parallel measurements were made three times, and the mean value was calculated.
3.4.6. Microstructure. The morphologies of different films were observed under a field emission scanning electron microscope (JSM-7001F, JEOL Ltd., Japan). A 5 mm × 5 mm film was fixed on the support using silver paste and observed at a magnification of ×5000. The cross-sections of the films were prepared by fracturing the films inside liquid nitrogen. All SEM samples were sputter-coated by gold particles before observation.

Fractionation of AHP
3.5.1. General Methods. The total sugar content was determined by the phenol-sulfuric acid assay using glucose as standard. Uronic acid contents were determined by the mhydroxydiphenyl method using galacturonic acid as a standard. Gel permeation and anion-exchange chromatographies were monitored by assaying the total sugar and uronic acid contents.

Analytical
Chromatography on DEAE-Cellulose. AHP (5 mg) was dissolved in distilled water (1 mL) and loaded on a DEAE-(diethylaminoethyl-) cellulose column (1:5 × 14 cm, Cl − ) pre-equilibrated with distilled water. The column was eluted first with 50 mL distilled water at 1.0 mL/min and then with a linear gradient from 0 to 1.0 M NaCl within 300 mL. The eluate was collected at 4 mL per tube and assayed for total sugar and uronic acid contents.   [42]. It was carried out at 40°C using a TSK-GEL G-4000PWxl column (7:8 mm × 300 mm, TOSOH, Japan) connected to a Shimadzu HPLC system. The column was precalibrated with standard dextrans (410 kDa, 150 kDa, 50 kDa and 5 kDa), and the elution volumes were plotted against the logarithm of their respective molecular weights. 20 μL sample (5 mg/mL) was injected, eluted with 0.2 M NaCl at a flow rate of 0.6 mL/min and monitored using a refractive index RID-10A detector (Shimadzu, Tokyo, Japan). The elution volume of pectin was plotted in the same graph, and the Mw was determined.
3.6.2. DM Determination of AHP. The degree of methyl esterification was determined by the standard titration method [46]. 0.5 g of pectin was dissolved in 50 mL distilled water and stirred for 10 min with 15 mL mixture reagent (5 mL concentrated hydrochloric acid with 100 mL 60% ethanol). The liquid was filtrated by the core funnel and washed six times with 15 mL mixture reagent above. Then, 60% ethanol was used to wash until no chloridion left in the liquid. The filter cake was dried for 1 h at 105°C. 60-80 mg samples were weighed and dissolved in 100 mL distilled water with two drops of phenolphthalein indicator. 0.02 mol/L NaOH solution was titrated until the phenolphthalein indicator become red, and the volume of NaOH solution was recorded (V 1 ). After titration, 20 mL 0.5 mol/L NaOH solution was added  9 International Journal of Polymer Science and shaken fiercely for 15 min. Equimolar 0.5 mol/L HCl solution was added with two drops of phenolphthalein indicator and was shaken sufficiently. 0.02 mol/L NaOH solution was used to titrate, and the volume of NaOH solution was recorded (V 2 ). The degree of methyl esterification in pectin was obtained by calculating V 2 /(V 1 + V 2 ).
3.6.4. FT-IR Analysis. Fourier transform infrared spectra were obtained on a Nicolet 6700 FT-IR spectrometer (Thermo Scientific, USA) with a DTGS detector in a range of 4000-400 cm -1 . The sample was measured as a film on KBr discs.
3.6.5. NMR Spectroscopy. The NMR experiments were conducted as described by Yan [47]. The samples (20 mg) were dissolved in D 2 O (99.8%, 0.5 mL), freeze-dried, and redissolved in D 2 O. The spectra were recorded at 25°C after 57000 scans on a Bruker Avanced 600 MHz spectrometer (Germany), operated at 150 MHz.
3.7. Statistical Analysis. The experiment was duplicated with three observation numbers adapted for each experiment. Analysis of variance (ANOVA) was performed with raw data, and mean values and the standard errors were calculated by the Statistical Analysis System (12.5.192.7). Differences among the mean values were determined by Duncan's multiple range test, and results with p < 0:05 were considered statistically significance.

Conclusions
In this study, the film-forming characterization and molecular structure of AHP extracted from deseeded sunflower heads have been investigated. The HPMC/AHP blended solutions showed better film-forming performance than that of pure HPMC on the aspect of viscosity, loss on drying, water vapor permeability and disintegration. AHP was also separated and purified into one neutral fraction and two acidic fractions (AHPA-1 and AHPA-2). Structure analysis showed that AHP had high content of the HG domain with low DM and high molecular weights, so AHP was speculated to have excellent gel properties. The addition of AHP could increase the brittleness of films by forming gel crosslinking structure and decrease tensile strength by making the blends more immiscible. Therefore, the natural low DM pectin from sunflower heads was satisfactory for film forming and had a good prospect as capsule ingredients in plant hollow capsule industries. It not only solved many problems having been unveiled in gelatin capsules but also improved the recycling and reusing of sunflower heads.

Data Availability
No data were used to support this study.

Disclosure
The sponsors had no role in the design, execution, interpretation, or writing of the study.

Conflicts of Interest
The authors declare no conflict of interest.