Hydrogels with or without chemical cross-linking have been studied and used for biomedical applications, such as tissue repair, surgical sealants, and three dimensional biofabrication. These materials often undergo a physical sol–gel or gel–sol transition between room and body temperatures and can also be chemically cross-linked at these temperatures to give dimensional stability. However, few studies have clearly shown the effect of heating/cooling rates on such transitions. Moreover, only a little is known about the effect of cross-linking temperature or the state on the modulus after cross-linking. We have established rheological methods to study these effects, an approach to determine transition temperatures, and a method to prevent sample drying during measurements. All the rheological measurements were performed minimizing the normal stress build-up to compensate for the shrinking and expansion due to temperature and phase changes. We chemically modified gelatin to give gelatin methacryloyl and determined the degree of methacryloylation by proton nuclear magnetic resonance. Using the gelatin methacryloyl as an example, we have found that the gel state or lower temperature can give more rigid gelatin-based polymers by cross-linking under visible light than the sol state or higher temperature. These methods and results can guide researchers to perform appropriate studies on material design and map applications, such as the optimal operating temperature of hydrogels for biomedical applications. We have also found that gelation temperatures strongly depend on the cooling rate, while solation temperatures are independent of the heating rate.

Various physical hydrogels and their chemically cross-linked materials1–8 have been studied for biomedical applications, including tissue repair, surgical sealants, and three dimensional (3D) biofabrication. Their importance and applications are actively growing in a wide range of fields. Some of these types of materials can undergo a physical sol–gel or gel–sol transition between room and body temperatures. Understanding the transitions and transition temperatures of hydrogels is crucial to their final applications since their mechanical properties can be dramatically different at these temperatures; for example, gelatin-based materials are sol at body temperature and gel at room temperature.9 

The rheological properties are bulk properties that greatly affect the flow properties during processes, the dimensional suitability of those materials, and their functions in applications. Hence, it is often used not only to characterize and identify the materials, as those properties are very sensitive to molecular structures, but also to evaluate and screen novel materials to target applications because flow and mechanical properties are identifying characteristics of the materials. Thus, proper rheometric characterization of the materials and the evaluation of the results are highly important.10 To study sol–gel and gel–sol transitions and their kinetics and to identify the transition temperatures, rheometry, such as temperature sweep tests,9 is often employed. However, these behaviors are usually studied at single cooling and heating rates even though the kinetics and transition temperatures can be strong functions of the rates. Thus, only a few studies have clearly shown the effect of heating/cooling rates on sol–gel and gel–sol transitions. A lack of data on the effect of the rate may mislead further studies, for example, trying to form a physical gel at an inappropriate temperature.

The current trend of studying tissue engineering using hydrogels is tuning of mechanical stiffness of the materials from many different angles.11 One of those is functionalizing materials for hydrogels to be chemically cross-linked to provide better dimensional stability.12 The effect of temperature or state (gel or sol) during cross-linking on the physical properties of the cross-linked materials is not well known even though the effect should be known prior to screening materials for targeting applications. In this work, as a representative compound, we chose gelatin methacryloyl (GEL-MA), a compound increasingly utilized in biomedical applications.12,13 Van Vlierberghe et al.,8 Yue et al.,14 and Klotz et al.15 reviewed this material in depth, including its synthesis and applications. Rizwan et al.16 determined G′ of GEL-MA after cross-linking at 4 °C and 37 °C. They showed that G′ of the sample cross-linked at 4 °C is much higher than the one cross-linked at 37 °C. Rebers et al.17 and Schuurman et al.18 also found that the compressive stress of GEL-MA cross-linked at a low temperature was higher than that cross-linked at 37 °C. However, both groups incubated their samples in a buffer solution after cross-linking prior to loading to their instrument for mechanical measurements. Thus, other variables or artifacts such as thermal history, drying, sample loading issues, and the degree of swelling may have been introduced.

We established careful, rheological methods to study the effects of cooling and heating rates on sol–gel and gel–sol transitions, as well as a method to prevent significant sample drying during measurements. We used an in situ protocol (i.e., rheological characterization was performed as the sample was cross-linked) imposing more controlled environment in order to minimize such additional variables to examine the optimal state or temperature that produces more rigid gels by cross-linking hydrogels under visible light. It was confirmed that cross-linking chemically modified gelatins (GEL-MA) at low temperatures can yield higher modulus (strength) than that cross-linked at a high temperature. The methods and results of this study can guide other researchers to perform comparable studies to design and screen specific materials and/or to map applications, including the determination of temperature windows of hydrogels for biomedical applications.

Deionized and purified (DI) water was generated by Millipore Milli-Q® Integral 5 Water Purification System. Deuterium oxide (D2O, 99.994 at. % D) for nuclear magnetic resonance measurement was purchased from Sigma-Aldrich (USA). As a base material, type A gelatin (from porcine skin with 300 bloom strength) was obtained from Sigma-Aldrich (USA). CorningTM Cell Culture Phosphate Buffered Saline with concentration 1X (1X PBS) was purchased from Fisher Scientific (USA). Carbonate–bicarbonate buffer of 0.25M and pH 9.6 was prepared by comprising 8.33 g of sodium carbonate (purity ≥99.5%, Fisher Chemical, USA) and 14.4 g of sodium bicarbonate (purity ≥99.7%, Fisher Chemical, USA) in 1 L DI water. To purify gelatin methacryloyl, the dialysis membrane (SnakeskinTM Dialysis Tubing, 10 k molecular weight cut off, Thermo Scientific, USA) was used. Methacrylic anhydride (MAA, purity ≥94%) was supplied by Sigma-Aldrich (USA). HoneywellTM sodium hydroxide solution (5N NaOH) was obtained from Fischer Scientific (USA) to control pH of solutions. The photo-initiator system is composed of eosin Y (certified biological stain, die content ≥88%, Fisher Chemical, USA), triethanolamine (gas chromatography grade, purity ≥99%, Sigma-Aldrich, USA), and 1-vinyl-2 pyrrolidinone (containing sodium hydroxide as an inhibitor, purity ≥99%, Sigma-Aldrich, USA). As a drying prevention measure, TeflonTM-like perfluoropolyether (PFPE) oil (KrytoxTM general purpose oil 103, DuPont, USA) was used.

Gelatin A was chemically modified12 with MAA to give gelatin methacryloyl (GEL-MA, Fig. 1) for further cross-linking using the introduced double bonds to increase mechanical stability. A 20% (w/v) solution of the gelatin (GEL) was prepared in 0.25M carbonate–bicarbonate buffer. This solution was adjusted to pH 9 using 5N NaOH. Methacrylic anhydride was added dropwise to the reaction vessel at 50 °C while being stirred by a magnetic bar at 600 rpm to achieve a final reaction ratio of 0.2 ml MAA per gram of dried GEL. The reaction was allowed to proceed for 2 h prior to dilution with five volumes of 1X PBS and pH adjustment to 7.4. Mainly two methacryl groups can be formed: methacrylamide and methacrylate (Fig. 1). To remove unreacted reactants and byproducts, this solution was placed in the dialysis membrane tubing to be dialyzed against DI water, which was replaced with fresh one every day for 7 days to purify and was lyophilized (Labconco FreeZone Benchtop Freeze Dry System) for 3 days. More detailed procedure can be found elsewhere.12 The chemical modification of GEL to GEL-MA was verified by proton nuclear magnetic resonance (1H-NMR)12,19,20 (BRUKER ASCEND, 500 MHz). GEL and GEL-MA were dissolved in D2O at concentrations of 25 mg/ml. Chemical shifts were recorded relative to the D2O peak and a total of 32 scans were taken per sample.

FIG. 1.

Synthesis of gelatin methacryloyl (GEL-MA); mainly two methacryl groups can be formed: methacrylamide and methacrylate. Characters a through e indicate distinct protons for 1H-NMR analysis.

FIG. 1.

Synthesis of gelatin methacryloyl (GEL-MA); mainly two methacryl groups can be formed: methacrylamide and methacrylate. Characters a through e indicate distinct protons for 1H-NMR analysis.

Close modal

Lyophilized GEL-MA was dissolved in 1X PBS to give 35% (w/v) solution at 37 °C. All rheological measurements without cross-linking were performed with this solution. A visible-light photoinitiator system was added to 35% (w/v) GEL-MA/PBS solution at 37 °C in a darkroom at the final concentrations of 0.01 mM eosin Y photoinitiator, which has a highest absorbance peak at 510 nm (green), 7.5 mM triethanolamine, and 37.5 mM 1-vinyl-2 pyrrolidinone. All rheological tests for the effect of cross-linking were carried out after or during cross-linking this solution.

We used a rotational rheometer (DHR-3, TA Instruments), which is one of the most popular physical characterization instruments for studying hydrogels, with parallel plates (top: quartz plate of 50 mm diameter, bottom: Peltier plate) in order to carry out small-amplitude oscillatory shear tests and obtain storage and loss moduli (G′ and G″). To account for any temperature or phase changes involved, tests were carried out while the gap was controlled to minimize the normal stress build-up by setting the sensitivity of the normal force as 0.1N to reflect the shrinkage or expansion caused by temperature or phase changes. Otherwise, normal stress build-up will affect the results significantly. An enhanced rheometer inertia correction21 was not used since the required system was not available.

1. Sample loading and drying prevention

A protocol for sample loading was developed, in which all samples were preheated above the solation-end temperature (Tsol,end, please refer to Sec. III B 3) to make sure that the sample is at the sol state prior to loading to the rheometer. If the sample was at the gel state during loading, the final shape of the sample loaded would be irregular and the contact between the sample and plates would not be perfect.

During rotational rheometry, in which a sample is sandwiched by two plates, the sample edge is exposed to the environment and is subject to drying. Since the torque near the sample edge contributes to the total torque the most, drying on the edge quickly skews the results. There have been a few methods to prevent sample edge from drying, such as applying silicone oil,22 using an evaporation blocker,23 which may adversely result in absorption of more solvent or water into the sample during the experiments in case the chemical potential of the solvent in the environment is higher than that in the sample, applying water drops on the edge,16 which can lead to the similar issue mentioned above, or no prevention measures taken.24 It was found that placing a sample in a chamber with environmental, i.e., humidity control only slows down drying.25 Thus, a precise control of humidity around a sample to prevent this from drying is quite a challenging task.

In this study, a 1 ml sample was pre-heated to 37 °C or 40 °C and loaded to the final gap of 0.5 mm between two parallel plates at 37 °C or 40 °C. Right after loading, low-viscous PFPE oil was applied on the sample edge to prevent the sample from drying and humidity absorption. Since this type of fluorinated oil is inert, it does not interact with polymers26 or water. Oscillatory time sweep tests for GEL-MA without chemical cross-linking with 10 rad/s frequency (ω), 3% strain, and 0.5 mm gap, at 37 °C, were performed with (for 90 min) or without (for 15 min) PFPE oil applied on the sample edge to confirm effectiveness of drying prevention. While storage and loss moduli (G′ and G″) are determined, the gap was controlled to minimize the normal stress build-up to compensate the shrinking due to drying.

2. Temperature sweep test for physical gelation and solation

Temperature sweep tests with oscillatory shear were performed for GEL-MA without chemical cross-linking with 3.14 rad/s frequency and 10% strain, at four rates of 0.3 °C/min, 0.6 °C/min, 2.5 °C/min, and 10 °C/min after loading a sample at 40 °C. First, the sample was cooled down from 40 °C to 0 °C and then heated up to 40 °C while G′ and G″ are determined with the gap controlled to minimize the normal stress build-up to compensate the shrinking and expansion due to temperature and phase changes. Strain sweep tests were performed prior to temperature sweep tests to confirm that the rheological measurements at given strain and frequency were within the linear regime both at 0 °C and 40 °C. Three types of transition temperatures were obtained. The transition start temperature and end temperature of gelation and solation, respectively (Tgel,start, Tsol,start, Tgel,end, and Tsol,end) were determined from the plots of phase angle vs temperature where the deviation is more than 1% from the linear trend obtained from the first four points. The crossover transition temperatures (Tgel,x and Tsol,x) were obtained where G′ = G″, at which temperatures are often called gelation point27 and melting28 or solation point determined during cooling and heating, respectively. However, since there were no temperature points where the crossover occurred, the interpolated temperature using linear modeling of moduli vs temperature was used.

3. Chemical cross-linking and frequency sweep test

Chemical cross-linking was performed with 35% (w/v) solution of GEL-MA in 1X PBS with the visible-light photoinitiator system in the darkroom at two temperatures. A sample solution of 1 ml was loaded on the rheometer to a gap of 0.5 mm with quartz upper plate at 37 °C and was cross-linked either at this temperature or at 23 °C after being cooled down to 23 °C at 0.3 °C/min while the gap was controlled to minimize the normal stress build-up. Cross-liking was performed by turning on green LED lights above the quartz top plate for five minutes, and after turning off the lights, frequency sweep tests were performed between 0.1 rad/s and 100 rad/s and with 0.5% strain either at 23 °C or 37 °C. If the temperature of frequency sweep differed from the cross-linking temperature, the sample was heated up or cooled down at 0.3 °C/min to the final temperature while the gap was controlled to minimize the normal stress build-up. The resulting storage modulus, G′ was used to compare properties of gels cross-linked at different temperatures. Strain sweep tests after cross-linking were performed prior to the frequency sweep tests to confirm that the rheological measurements at 100 rad/s and 0.5% strain were within the linear regime both at 23 °C and 37 °C. If the data at 100 rad/s are in the linear regime, all other data at lower frequency should be linear, so that strain sweep tests were performed only at 100 rad/s.

Time sweep tests29 were also performed during chemical cross-linking with lights on. Since cross-linking kinetics is out of the scope of this project, the results are not shown. However, it was found that there was no increase in moduli after turning off the light during the time sweep tests and thus, we can infer that no cross-linking should have occurred during the frequency sweep tests without the light source.

Statistical analysis was performed by a one-side Student t-test using Microsoft Excel. Statistical significance was considered when the p-value was lower than 0.05. Two different significance levels were marked: * for p < 0.05 and ** for p < 0.01. The error bars in plots represent the standard deviation. The number of repeats are given in the relevant sections.

The chemical modification of GEL to GEL-MA was verified by 1H-NMR. Figure 2 shows the chemical modification of GEL to GEL-MA by 1H-NMR. We assigned the peaks, as Claaßen et al. reported.20 One of the vinyl methylene (double bound) hydrogens (a in Fig. 1) of methacrylate groups has a tiny peak at 5.9–6.3 ppm. The other hydrogen (b) of methacrylate groups and one of the double bond hydrogens (c) of methacrylamide groups have peaks at 5.55–5.75. The other hydrogen (d) of methacrylamide groups and γ-hydrogen (x) of methacryl-modified hydroxyproline have peaks at 5.35–5.5. Methyl hydrogens (e) of methacrylamide and methacrylate groups have a peak at 1.95 ppm. Figure 2 shows that GEL-MA has peaks of vinyl methylene at 5.35 ppm and 5.7 ppm, which were not observed for pure GEL, implying the successful methacryloylation. Even if there was remaining unreacted MAA, this dissolves in water in small amount and is easy to be removed so that it would be safe to assume that there was no trace. However, the byproduct of methacrylic acid could be an issue. Figure 2 shows no apparent peak at 5.35 ppm, at which neutralized methacrylic acid has a peak,20 so that there was no trace of methacrylic acid. There is a small peak at 6.2 ppm, implying that only a few methacrylate groups were formed. Methacrylate groups could appear more if a high molar excess of MAA was used.30,31

FIG. 2.

1H-NMR spectra of GEL and GEL-MA with assignment of vinyl methylene protons of methacrylate group (a: 6.1 ppm and b: 5.55–5.75 ppm), those of methacrylamide group (c: 5.55–5.75 ppm and d: 5.35–5.5 ppm), methyl protons of both methacryl groups (e: 1.95 ppm), aromatic protons of phenylalanine (7.3–7.34 ppm), ε-protons of lysine (3.0 ppm), and γ-protons of modified hydroxyproline (x: 5.35–5.5 ppm).

FIG. 2.

1H-NMR spectra of GEL and GEL-MA with assignment of vinyl methylene protons of methacrylate group (a: 6.1 ppm and b: 5.55–5.75 ppm), those of methacrylamide group (c: 5.55–5.75 ppm and d: 5.35–5.5 ppm), methyl protons of both methacryl groups (e: 1.95 ppm), aromatic protons of phenylalanine (7.3–7.34 ppm), ε-protons of lysine (3.0 ppm), and γ-protons of modified hydroxyproline (x: 5.35–5.5 ppm).

Close modal

The degree of methacryloylation (DoM) of GEL-MA can be defined as the number of methacryloyl (methacrylamide and methacrylate: b + c in Fig. 1) groups bound to GEL over the number of amine groups (lysine and hydroxylysine) of GEL. Since aromatic side groups (phenylalanine and tyrosine) do not participate in methacryloylation32 as can be seen in Fig. 2 (no difference in the first blue band), the former is normalized by the number of aromatic side groups of GEL-MA, and the latter was normalized by that of GEL. Even though synthesis of GEL-MA was developed based on the reaction of amine groups (lysine and hydoxylysine) with MAA9 through nucleophilic attack and deprotonation by the amine to form amides, other reactions can also occur with groups other than amine groups. For example, alcohols can react with acid anhydrides to form esters and carboxylic acids. Serine on GEL can react with MAA to form methacrylate. However, we obtained the DoM only based on ε-hydrogens of amines (lysine and hydroxylysine),33 which appear at 2.9–3.0 ppm,20 to avoid complex, full analysis. Aromatic hydrogens in tyrosine appear at 6.8 ppm and 7.1 ppm while those in phenylalanine appear at 7.3–7.34 ppm.20Figure 2 shows that there are only small peaks between 6.8 ppm and 7.1 ppm, implying that there are much fewer tyrosine than phenylalanine on GEL, as Claaßen et al. reported.20 Thus, only the peaks around 7.3 (i.e., aromatic hydrogens in phenylalanine) are used to calculate DoM as shown in the following equation:

DoMmethacryloylGELMAphenylalanineGELMA÷lysine+hydroxylysineGELphenylalanineGEL,
(1)

where indicates the area under the peak for a specific side group. A DoM of 0.60 was obtained.

1. Drying prevention

Drying or evaporation of polymeric solution has a key role in most solution processes,34 especially when melt processes are not possible, or wet conditions are required. However, drying is one of the main causes of artifacts in characterization for the solutions or colloids. In this study, we confirmed the performance of our counter-measure for drying during characterization by rheometry. Oscillatory time sweep tests with 10 rad/s frequency, 3% strain, and 0.5 mm gap, at 37 °C, were performed with (for 90 min) or without (for 15 min) PFPE oil applied on the sample edge to confirm effectiveness of drying prevention. Figure 3 compares the storage modulus, G′ as a function of time with and without applying PFPE oil on the sample edge. The early values of the storage modulus for both cases are about the same, but after one minute, the difference becomes dramatic. The increase in G′ can be fitted to a linear equation

G(Pa)=0.0013t(min)+0.422,
(2)

for the case with oil and an exponential equation, where t is time in minutes.

G(Pa)=e0.62+0.51tmin,
(3)

for the case without drying prevention. While it takes 60 min for G′ of the sample with the applied oil to increase by 20%, that of the sample without the oil increases by 3000 times in 15 min. The dried edge of the sample without drying prevention could form a cap to slow down or prevent drying, but the results show that the drying process continues. Water molecules from inner sections of the sample should have continued diffusing through the dried edge and evaporated. Since we continually sheared the sample for the measurements, such diffusion could be enhanced, taking into account that diffusion is affected in a shear field.26,35 The edge of the sample without the drying prevention was not apparent visually after the test, but it could be isolated and detached from the sample with a soft yet elastic texture. The gap value was decreasing a lot while keeping the normal force within the tolerance of 0.1 N. This can be another indication of drying. The rheological behavior should be in the non-linear regime36,37 considering the magnitude of G′ at longer time without drying prevention, so that the definition of this variable may not be valid. However, the physical meaning of G′, the stiffness of the sample, would not change, and G′ can still show how an artifact (e.g., drying) can easily mislead research data. Meanwhile, the edge of the sample with the drying prevention did not show such a dramatic change, so it was not possible to identify the drying issue. Considering the low values of G′ at the sol state and its huge increases due to physical gelation (Fig. 4) and cross-linking, the 20% increase can be deemed negligible. Prevention of drying is very important, especially for chemical cross-linking because if drying plays a non-trivial role, it will be difficult to isolate the effect of cross-linking from the results. Moreover, it is known that drying can induce the formation of a certain structure, such as self-assembled cylinders38 or clustering of the slender microstructures,39 resulting in a more complex situation. The coffee ring effect40–42 can also be an issue since it can induce highly inhomogeneous distribution of additives. It should be noted that because the edge contributes most significantly to the torque measurements, the drying edge, which starts drying in parallel plates or cone–plate fixture in rotational rheometry, significantly affects the results even though the dried edge can positively serve as a barrier against further drying at longer time duration. It should also be noted that slow drying can still progress even with the drying prevention, so that the effect of drying needs to be considered, and the maximum test time should be identified prior to main experiments. For example, if the maximum change of 20% is allowed, a test even with the drying prevention should be completed in 60 min. The viscosity (0.1 Pa s) of PFPE oil should be lower than that of the samples so that the oil contributes little to the measured torque as shown in Fig. 3. Due to the hydrophobicity of the oil and interfacial slip, dragging of the sample43 would be lower than thought.

FIG. 3.

Time sweep tests with and without drying prevention, n = 1.

FIG. 3.

Time sweep tests with and without drying prevention, n = 1.

Close modal
FIG. 4.

Temperature sweep tests for physical gelation and solation of GEL-MA at a frequency of 3.14 rad/s and a strain of 10%, at four rates of 0.3 °C/min, 0.6 °C/min, 2.5 °C/min, and 10 °C/min in the darkroom: first, (a) cooling from 40 °C to 0 °C and then (b) heating back to 40 °C. Solid symbols and solid trend lines: G′, open symbols and dotted trend lines: G″, brown square: 0.3 °C/min, blue triangle: 0.6 °C/min, red diamond: 2.5 °C/min, and green circles: 10 °C/min. Mean values are depicted; to avoid confusion, error bars are not shown. The standard deviation of three separate tests with a fresh sample under each condition is lower than 20%.

FIG. 4.

Temperature sweep tests for physical gelation and solation of GEL-MA at a frequency of 3.14 rad/s and a strain of 10%, at four rates of 0.3 °C/min, 0.6 °C/min, 2.5 °C/min, and 10 °C/min in the darkroom: first, (a) cooling from 40 °C to 0 °C and then (b) heating back to 40 °C. Solid symbols and solid trend lines: G′, open symbols and dotted trend lines: G″, brown square: 0.3 °C/min, blue triangle: 0.6 °C/min, red diamond: 2.5 °C/min, and green circles: 10 °C/min. Mean values are depicted; to avoid confusion, error bars are not shown. The standard deviation of three separate tests with a fresh sample under each condition is lower than 20%.

Close modal

2. Effect of cooling and heating rates on phase transition

It is important to determine the sol–gel and gel–sol transition temperatures to screen or select proper materials for targeting applications. Temperature sweep tests are often used for this purpose, but most tests to date have been carried out at single cooling and heating rates9,19 without critical gap compensations for the volume change of the sample during the measurements. Considering that such transition temperatures are a strong function of the rates, it would be necessary to perform temperature sweep tests at various rates. Figure 4(a) shows the results of temperature sweep tests at four cooling rates from 40 °C to 0 °C. Gelatin-based polymers in water have a self-assembly property upon temperature change that is similar to many other biopolymers. This combination of GEL-MA and water shows an upper critical solution temperature (UCST), i.e., they form a one-phase solution at a higher temperature, but a two-phase colloid at a lower temperature due to hydrogen bonding, resulting in a 3-dimensional structure.8,44,45 The GEL-MA was at the sol state showing that the storage modulus is much smaller than the loss modulus (G′ ≪ G″) above 32 °C. Furthermore, cooling induces network formation (physical gelation)46 showing that G′ ≫ G″ and that G′ increases by 1.6 × 106 times while G″ increases only by 800 times at 0 °C compared to 40 °C, which implies that an almost pure solid-like structure (elastomer) has been formed regardless of the cooling rates. Figure 4(a) also shows the effect of the cooling rate on physical gelation. The higher the cooling rates are, the lower the temperatures at which the increases of G′ and G″ occur at, and the lower the gelation temperatures are, as shown in Fig. 5(a). During fast cooling, the specific structure did not form even when the temperature was sufficiently low. Meanwhile, as the temperature further decreases, the structure seems to appear at a temperature lower than the actual transition temperature. The cooling/heating mechanism of the rheometer by a Peltier bottom plate increases this effect since the sample is cooled down or heated up only on its bottom. Because it is not realistic to carry out temperature sweep tests at very low rates near zero to obtain rate-invariant behavior, the effect of the cooling rate on the gelation behavior should be taken into account. Thus, temperature sweep tests need to be performed at various rates, or the rate should be identified relevant to the actual application. Figure 4(b) shows the results of temperature sweep tests from 0 °C to 40 °C at four heating rates. Heat destroys the network, resulting in dramatic decreases in moduli back to the original values at the sol state. All the curves show similar behavior, indicating that the heating rate has little effect on the physical solation and solation temperatures [Fig. 5(b)]. Such a difference in terms of kinetics between cooling and heating originates from structure construction and destruction. Longer time is required to construct than to destruct certain structures, such as networks. Park et al.47 showed a similar trend for the melting and crystallization of polyolefins. One may argue that viscous heating could affect the results, but we can reason that the magnitude of such an effect was small in this small-amplitude oscillatory shear flow unlike the significant effect of viscous heating in large-amplitude oscillatory shear flow48 or steady simple flow.49 If it was significant, we should have seen differences among heating curves since different time periods of each rate should induce different heat dissipation.

FIG. 5.

Phase angle vs temperature of temperature sweep tests for GEL-MA without cross-linking at a frequency of 3.14 rad/s and a strain of 10%. (a) Phase angle vs temperature during cooling at 0.3 °C/min, (b) phase angle vs temperature during heating at 0.3 °C/min. Temperatures deviating from the linear trend are defined as gelation or solation start temperature and end temperature (Tgel,start, Tsol,start, Tgel,end, and Tsol,end). The blue and red curved arrows show the order of measurements. Three separate tests with a fresh sample were performed, but results of only one test were shown for the demonstration purpose.

FIG. 5.

Phase angle vs temperature of temperature sweep tests for GEL-MA without cross-linking at a frequency of 3.14 rad/s and a strain of 10%. (a) Phase angle vs temperature during cooling at 0.3 °C/min, (b) phase angle vs temperature during heating at 0.3 °C/min. Temperatures deviating from the linear trend are defined as gelation or solation start temperature and end temperature (Tgel,start, Tsol,start, Tgel,end, and Tsol,end). The blue and red curved arrows show the order of measurements. Three separate tests with a fresh sample were performed, but results of only one test were shown for the demonstration purpose.

Close modal

3. Effect of cooling and heating rates on transition temperatures

The transition start temperature and end temperature (Tgel,start, Tsol,start, Tgel,end, and Tsol,end) were obtained from the plots of phase angle vs temperature to give physical significances of transition starting and completing as shown in Fig. 5.

Figure 6 shows three different types of transition temperatures for gelation and solation, respectively, transition start and end temperatures (Tgel,start, Tsol,start, Tgel,end, and Tsol,end) and crossover transition temperatures (Tgel,x, Tsol,x). The heating rate has little effect on all of the transition temperatures, as mentioned earlier, but the transition temperatures during cooling decrease with the cooling rate. The area between Tgel,start and Tgel,end in Fig. 6(a) during cooling and that between Tsol,start and Tsol,end in Fig. 6(b) during heating represent the mixed states of sol and gel. Thus, it would be better to avoid these areas in applications requiring either the sol or the gel state.

FIG. 6.

Gelation and solation temperatures of GEL-MA. (a) Gelation temperatures determined during cooling from 40 °C to 0 °C and (b) solation temperatures determined during heating from 0 °C to 40 °C. Data are presented as means with the standard deviation for three separate tests with a fresh sample under each condition. Error bars indicate standard deviations of n = 3 and * for p < 0.05 and ** for p < 0.01.

FIG. 6.

Gelation and solation temperatures of GEL-MA. (a) Gelation temperatures determined during cooling from 40 °C to 0 °C and (b) solation temperatures determined during heating from 0 °C to 40 °C. Data are presented as means with the standard deviation for three separate tests with a fresh sample under each condition. Error bars indicate standard deviations of n = 3 and * for p < 0.05 and ** for p < 0.01.

Close modal

The effect of the cooling rate on the gelation temperatures was modeled as follows:

Tgel,start=1.41 ln CR+28.2(°C),
(4)
Tgel,x=1.38 ln CR+26.0(°C),
(5)
Tgel,end=1.66 ln CR+20.5(°C),
(6)

where CR is the cooling rate in °C/min. Since there was no statistical difference in solation temperatures among different heating rates, the average was taken to give Tsol,start = 26.3 °C, Tsol,x = 30.7 °C, and Tsol,end = 33.5 °C. Sewald et al.28 reported 14.5 °C ≤ Tgel,x ≤ 23.6 °C and 21.8 °C ≤ Tsol,x ≤ 29.9 °C determined at cooling/heating rates of 1 °C/min, a strain amplitude of 5%, and a frequency of 1 s−1 for their 10% w/w GEL-MA solutions with the various DoM. Our values estimated at 1 °C/min using above equations are Tgel,x = 26.0 °C and Tsol,x = 30.7 °C. Their method to determine the DoM is different from our method, so a direct comparison may not be possible, but we can still see the difference that our transition temperatures are higher than theirs. The difference may arise from the concentration of GEL-MA. The concentration of our GEL-MA (35%) is much higher than theirs (10%) and leads to the earlier physical gelation (i.e., easy to form the gel structure) up on cooling and later solation (i.e., difficult to break the gel structure) up on heating.

When a specific application requires the sol state of the sample, such as flow in a nozzle of a 3D bio-printer50 or sample loading to a rheometer, the process should be designed to operate above the solation temperature. Before applying the material to form a certain shape, the material should be pre-heated in the applications. If the sample has been stored at room temperature or below, it is at the gel state since solation can occur only above Tsol,start = 26 °C [Fig. 6(b)]. Furthermore, the sample should be heated above Tsol,end = 34 °C to complete the solation process and to have flow without any physical network. Otherwise, flow instability will lead to irregular shapes in the final products, which can be often seen in 3D bio-printing. For example, the temperature of the sample inside the barrel and nozzle zones of the 3D bio-printer should remain above Tsol,end. We may have considered a gelation temperature, especially Tgel,start, to set the nozzle temperature, i.e., maintaining a temperature above Tgel,start = 30 °C (at 0.3 °C/min). However, the value could be misleading because it is rate-dependent and may have been underestimated. As mentioned earlier, prior to sample loading into the rheometer, the sample was pre-heated to 40 °C; loading a partial gel on the rheometer plate can lead to poor contact between the rheometer plates and the sample causing serious measurement errors. Although gelation temperatures depend on the cooling rate, they still have a certain role in applications that involve cooling or physical gelation. If a physical gel is an efficient state prior to chemical cross-linking (this application will be discussed in Sec. III B 4), for example in 3D bio-printing, the sol, which has been heated above Tsol,end = 34 °C, should be applied on to a surface at a temperature below Tgel,end = 23 °C (at 0.3 °C/min). It should be noted that a physical gel is not the only possible state for 3D printing. For example, when a material with a high yield stress is used in 3D printing, we may expect that the printed material can support its structure even without a physical gel. However, more pressure should be applied to make the flow of the material in a syringe or barrel. Considering that the current trend of 3D-bioprinting improvement is focused on increasing the resolution, it would be better to have a range of possible pressures to be applied, especially toward low pressure. Thus, we are envisioning that a solid structure by gelation and cross-linking can be formed after printing with easy and fine flow with minimal pressure in 3D printing. If the cooling rate of the sample after applying it to the target surface is assumed to be slower than 0.3 °C/min, 23 °C is a safe indicator that the pure gel state has been reached. If the cooling rate of the sample on the target surface is assumed to be 10 °C/min, the surface should be as cold as 17 °C. Such sample cooling rates on the cold surface can be estimated based on the material properties. The crossover temperatures (Tgel,x and Tsol,x) are not practically useful since these values do not guarantee a pure sol or gel state, other than as a single variable to compare different materials. It should be noted that many other biomaterials show a lower critical solution temperature (LCST), i.e., they form a one-phase solution at a lower temperature, but a two-phase colloid at a higher temperature due to gelation.51 In this case, all analyses should be approached in the other direction.

4. Effect of cross-linking temperature on modulus

Frequency sweep tests were performed at frequencies between 0.1 rad/s and 100 rad/s at 0.5% of strain, at 23 °C or 37 °C after cross-linking for 5 min with green light emitting diodes (LEDs) light, utilizing four conditions in total. All the materials show a rubbery plateau region, as demonstrated in Fig. 7, and a clear effect of the cross-linking temperature on the modulus of final samples was observed. When the sample was cross-linked at the gel state (23 °C), the moduli was greater than that resulting from cross-linking at the sol state (at 37 °C), increases by 2.3 times when measured at 23 °C and 8 times when measured at 37 °C. One could hypothesize that the restricted motion of the molecules in the gel state limits chemical cross-linking between molecules. However, it is more likely that the proximity between molecules plays a more important role in cross-linking, resulting in a higher degree of cross-linking. One of the current issues in the use of hydrogels for biomedical applications lies in finding ways to increase the modulus of the hydrogels because natural body tissues are much tougher than man-made hydrogels derived from biomaterials. A higher modulus can be achieved by cross-linking to make a pre-formed patch at room temperature (G′ ≅ 21 kPa) prior to applying the material at body temperature (G′ ≅ 16 kPa) rather than applying a sol to the body and then cross-linking on the body (G′ ≅ 1.7 kPa).

FIG. 7.

Storage modulus of frequency sweep tests for GEL-MA determined at a strain of 0.5%, at 23 °C or 37 °C, which are shown in larger font and after chemical cross-linking at 23 °C or 37 °C, which are shown in smaller font. Each test was performed with a fresh new sample. Data represent means of triplicate determinations in one of the four similar replicate studies. Standard deviations are not shown, but were ≤10% for each triplicate mean data point.

FIG. 7.

Storage modulus of frequency sweep tests for GEL-MA determined at a strain of 0.5%, at 23 °C or 37 °C, which are shown in larger font and after chemical cross-linking at 23 °C or 37 °C, which are shown in smaller font. Each test was performed with a fresh new sample. Data represent means of triplicate determinations in one of the four similar replicate studies. Standard deviations are not shown, but were ≤10% for each triplicate mean data point.

Close modal

As mentioned in the Introduction, it was found that cross-linking at a low temperature can yield more rigid hydrogels.16–18 Two major differences are that we used an in situ protocol (i.e., rheological characterization was performed as the sample was cross-linked), imposing more controlled environment to minimize additional variables, such as swelling and thermal history and that we did not swell the cross-linked GEL-MA. Our results show that G′ of cross-linked GEL-MA is a function of temperature, while others16,18,28 showed that the stiffness of their cross-linked GEL-MA was almost independent of the temperature. The storage modulus, G′(ω), in the plateau region, including the plateau modulus is known to be a weak function of the density and temperature.52 Thus, we should expect that the modulus changes with the temperature, but the degree of the change can be different for different cases. They made their sample swollen after cross-linking, while we dissolved GEM-MA first and then cross-linked. These two cases should give different material properties.53 The exact cause to the difference would be investigated further, but we can consider three possible reasons: flow history,54 the degree of swelling, and the degree of cross-linking. In case the polymer network is fully swollen by water, its properties may depend on the temperature little since the properties of water change little within the temperature window. In case the polymer is not fully cross-linked, the portion of the physical gel and sol will lead to the strong dependence of properties on the temperature. As a result, the properties of their material changed within the plateau region, where the effect of temperature is little, in the temperature window, while those of our material moved from the plateau region to the flow region by increasing the temperature, so that the modulus shows a lower value at a higher temperature.

Figure 7 can be used to select materials or to guide their applications. For instance, if a more rigid medical sealant derived from the same material is needed, the material can be cross-linked at temperatures lower than its gelation temperature to form a patch, which can then be applied at room temperature [Fig. 8(a)], such as a dry fibrin sealant.55 When materials are developed for 3D bio-printed in vivo artificial organs11,56 or cell culture systems,57 the material can be selected after cross-linking at temperatures lower than its gelation temperature to form a rigid structure and would still have high mechanical strengths to be applied at body temperature [Fig. 8(b)]. On the contrary, if a liquid-type injectable hydrogel51,58 is desired, materials can be designed or selected to give the viscosity low enough to flow nicely at body temperature and to become rigid enough at body temperature after cross-linking at body temperature [Fig. 8(d)]. The GEL-MA at 35% w/v used in this study provides an insufficient modulus (G′ ≅ 1.7 kPa) for this purpose, thus higher concentrations or higher degrees of substitution of methacryl groups with more photo-initiators need to be utilized. If the targeted application operates at room temperature, the data at 23 °C need to be collected and examined. For example, mesh structures made of biomaterials for protein purification are of interest59 [Fig. 8(c)]. All above algorithms can be applied for materials similar to the GEL-MA studied in this project, and more research on other materials is necessary to see that the same algorithms can be used.

FIG. 8.

Different processing temperatures for different applications; both cross-linking and application temperatures.

FIG. 8.

Different processing temperatures for different applications; both cross-linking and application temperatures.

Close modal

The average molecular weight between two neighboring cross-linking points (Mc) is one of the most important parameters utilized to characterize the network structure.9,60 The network structure of hydrogels can be obtained61 by

Mc=ρRT/G,
(7)

where ρ is the density, R is the gas constant, and G is shear modulus, which is G′, where the phase angle is the minimum in this study. Even though this equation was developed for rubber elasticity, this should be valid for both solutions and melts.62 When cross-linked at 23 °C, Mc is ten times smaller than that when cross-linked at 37 °C (in case the measurements were performed at body temperature), implying that there are ten times more cross-linking points, resulting in a more rigid structure. Such cross-linking temperature should be selected based on Tgel,end, which was 23 °C (at 0.3 °C/min) for this study, as mentioned earlier, to ensure that the sample was in the gel state prior to cross-linking.

We have shown that rheometry can be employed to characterize hydrogels with or without cross-linking and to develop and select materials for biomedical applications. The temperature or state (gel or sol) during chemical cross-linking of hydrogels has an important role in applications because the final properties including modulus can vary significantly with the processing temperature. Cross-linking gelatin methacryloyl at the gel state can lead to much higher modulus than that at the sol state. Prior to chemical cross-linking, phase transition (gelation or solation) temperatures should be carefully identified. The temperature changing rates, especially the cooling rate, play an important role in such determination so that transition temperatures should be identified at a proper cooling rate. We used phase angles from rheometry to identify the transition temperatures. Three different transition temperatures (crossover, start, and end) can be utilized for each application to define where and when the materials should be used. To achieve more accurate and precise rheological data, samples should be prevented from drying during measurements by applying inert oil, such as low viscosity inert oil, on the sample edge.

Partial financial support from the Department of Chemical and Process Engineering, University of Canterbury (New Zealand), the Department of Mechanical and Industrial Engineering, University of Toronto (Canada), and the Department of Defense, Medical Discovery Award under Grant No. R01 HL127144-01 (USA) is acknowledged. We also thank Dr. Monika Ivancic at the University of Vermont, USA, for the advice on 1H-NMR analysis and Dr. Ying Wai Lam and Catrina Hood at the University of Vermont, USA, for the use of a lyophilizer. H.E.P. thanks Dr. Benefsha Mohammad at Danbury Hospital, USA, for advice on applying medical sealants during surgery.

1.
R.
Parhi
, “
Cross-linked hydrogel for pharmaceutical applications: A review
,”
Adv. Pharm. Bull.
7
(
4
),
515
530
(
2017
).
2.
J.
Michalek
,
M.
Pradny
,
K.
Dusek
,
M.
Duskova
, and
R.
Hobzova
,
Hydrogel in Biology and Medicine
(
Nova Science Publishers
,
New York
,
2010
).
3.
S.
Rimmer
,
Biomedical Hydrogels; Biochemistry, Manufacture and Medical Applications
(
Woodhead Publishing
,
Oxford
,
2011
).
4.
R. M.
Ottenbrite
,
K.
Park
, and
T.
Okano
,
Biomedical Applications of Hydrogels Handbook
(
Springer
,
New York
,
2010
).
5.
R.
Barbucci
,
Hydrogels; Biological Properties and Applications
(
Springer
,
Milan
,
2009
).
6.
S.
Kalia
,
Polymeric Hydrogels as Smart Biomaterials
(
Springer
,
Cham
,
2016
).
7.
D.
DeRossi
,
K.
Kajiwara
,
Y.
Osada
, and
A.
Yamauchi
,
Polymer Gels; Fundamentals and Biomedical Applications
(
Plenum Press
,
New York
,
1989
).
8.
S.
Van Vlierberghe
,
P.
Dubruel
, and
E.
Schacht
, “
Biopolymer-based hydrogels as scaffolds for tissue engineering applications: A review
,”
Biomacromolecules
12
,
1387
1408
(
2011
).
9.
A. I. V. D.
Bulcke
,
B.
Bogdanov
,
N. D.
Rooze
,
E. H.
Schacht
,
M.
Cornelissen
, and
H.
Berghmans
, “
Structural and rheological properties of methacrylamide modified gelatin hydrogels
,”
Biomacromolecules
1
,
31
38
(
2000
).
10.
S. B.
Ross-Murphy
, “
Structure and rheology of gelatin gels: recent progress
,”
polymer
33
(
12
),
2622
2627
(
1992
).
11.
C.
Yu
,
W.
Zhu
,
B.
Sun
,
D.
Mei
,
M.
Gou
, and
S.
Chen
, “
Modulating physical, chemical, and biological properties in 3D printing for tissue engineering applications
,”
Appl. Phys. Rev.
5
,
041107
(
2018
).
12.
H.
Shirahama
,
B. H.
Lee
,
L. P.
Tan
, and
N.-J.
Cho
, “
Precise tuning of facile one-pot gelatin methacryloyl (GelMA) synthesis
,”
Sci. Rep.
6
,
31036
(
2016
).
13.
Y.
Chen
,
R.
Lin
,
H.
Qi
,
Y.
Yang
,
H.
Bae
,
J. M.
Melero-Martin
, and
A.
Khademhosseini
, “
Functional human vascular network generated in photocrosslinkable gelatin methacrylate hydrogels
,”
Adv. Funct. Mater.
22
,
2027
2039
(
2012
).
14.
K.
Yue
,
G. T.-d.
Santiago
,
M. M.
Alvarez
,
A.
Tamayol
,
N.
Annabi
, and
A.
Khademhosseini
, “
Synthesis, properties, and biomedical applications of gelatin methacryloyl (GelMA) hydrogels
,”
Biomaterials
73
,
254
271
(
2015
).
15.
B. J.
Klotz
,
D.
Gawlitta
,
A. J.
Rosenberg
,
J.
Malda
, and
F. P.
Melchels
, “
Gelatin-methacryloyl hydrogels: Towards biofabrication-based tissue repair
,”
Trends Biotechnol.
34
(
5
),
394
407
(
2016
).
16.
M.
Rizwan
,
G. S.
Peh
,
H.-P.
Ang
,
N. C.
Lwin
,
K.
Adnan
,
J. S.
Mehta
,
W. S.
Tan
, and
E. K.
Yim
, “
Sequentially-crosslinked bioactive hydrogels as nano-patterned substrates with customizable stiffness and degradation for corneal tissue engineering applications
,”
Biomaterials
120
,
139
154
(
2017
).
17.
L.
Rebers
,
T.
Granse
,
G. E.
Tovar
,
A.
Southan
, and
K.
Borchers
, “
Physical interactions strengthen chemical gelatin methacryloyl gels
,”
Gels
5
(
1
),
4
(
2019
).
18.
W.
Schuurman
,
A.
Levett
,
M. W.
Pot
,
R.
van Weeren
,
W. J. A.
Dhert
,
D. W.
Hutmacher
,
F. P. W.
Melchels
,
T. J.
Klein
, and
J.
Malda
, “
Gelatin-methacrylamide hydrogels as potential biomaterials for fabrication of tissue-engineered cartilage constructs
,”
Macromol. Biosci.
13
,
551
561
(
2013
).
19.
B. H.
Lee
,
N.
Lum
,
L. Y.
Seow
,
Q.
Lim
, and
L. P.
Tan
, “
Synthesis and characterization of types A and B gelatin methacryloyl for bioink applications
,”
Materials
9
,
797
(
2016
).
20.
C.
Claaßen
,
M. H.
Claaßen
,
V.
Truffault
,
L.
Sewald
,
G. E. M.
Tovar
,
K.
Borchers
, and
A.
Southan
, “
Quantification of substitution of gelatin methacryloyl: Best practice and current pitfalls
,”
Biomacromolecules
19
,
42
52
(
2018
).
21.
R.
Hudson
,
A.
Holder
,
K.
Hawkins
,
P. R.
Williams
, and
D.
Curtis
, “
An enhanced rheometer inertia correction procedure (ERIC) for the study of gelling systems using combined motor-transducer rheometers
,”
Phys. Fluids
29
,
121602
(
2017
).
22.
K. A.
Vakalopoulos
,
Z.
Wu
,
L.
Kroese
,
G.-J.
Kleinrensink
,
J.
Jeekel
,
R.
Vendamme
,
D.
Dodou
, and
J. F.
Lange
, “
Mechanical strength and rheological properties of tissue adhesives with regard to colorectal anastomosis
,”
Ann. Surg.
261
(
2
),
323
(
2015
).
23.
J.
Sato
and
V.
Breedveld
, “
Evaporation blocker for cone-plate rheometry of volatilee samples
,”
Appl. Rheol.
15
,
390
397
(
2005
).
24.
R. A.
Pouliot
,
A.
Link
,
N. S.
Mikhaiel
,
M. B.
Schneck
,
M. S.
Valentine
,
F. J. K.
Gninzeko
,
J. A.
Herbert
,
M.
Sakagami
, and
R.
Heise
, “
Development and characterization of a naturally derived lung extracellular matrix hydrogel
,”
J. Biomed. Mater. Res., Part A
104
,
1922
1935
(
2016
).
25.
P.
Aprelev
,
B.
McKinney
,
C.
Walls
, and
K. G.
Kornev
, “
Magnetic stage with environmental control for optical microscopy and high-speed nano- and microrheology
,”
Phys. Fluids
29
,
072001
(
2017
).
26.
H. E.
Park
and
J. M.
Dealy
, “
Effects of pressure and supercritical fluids on the viscosity of polyethylene
,”
Macromolecules
39
,
5438
5452
(
2006
).
27.
H. H.
Winter
, “
Gel point
,” in
Encyclopedia of Polymer Science and Technology
, 4th ed. (
John Wiley & Sons
,
2016
).
28.
L.
Sewald
,
C.
Claaßen
,
T.
Götz
,
M. H.
Claaßen
,
V.
Truffault
,
G. E. M.
Tovar
,
A.
Southan
, and
K.
Borchers
, “
Beyond the modification degree: Impact of raw material on physicochemical properties of gelatin type A and type B methacryloyls
,”
Macromol. Biosci.
18
,
1800168
(
2018
).
29.
D.
Khandal
,
B.
Riedl
,
J. R.
Tavares
,
J.
Carreau
, and
M.-C.
Heuzey
, “
Tailoring cellulose nanocrystals rheological behavior in aqueous suspensions through surface functionalization with polyethyleneimine
,”
Phys. Fluids
31
,
021207
(
2019
).
30.
E.
Hoch
,
C.
Schuh
,
T.
Hirth
,
G. E. M.
Tovar
, and
K.
Borchers
, “
Stiff gelatin hydrogels can be photo-chemically synthesized from low viscous gelatin solutions using molecularly functionalized gelatin with a high degree of methacrylation
,”
J. Mater. Sci.: Mater. Med.
23
,
2607
2617
(
2012
).
31.
B. H.
Lee
,
H.
Shirahama
,
N.-J.
Cho
, and
L. P.
Tan
, “
Efficient and controllable synthesis of highly substituted gelatin methacrylamide for mechanically stiff hydrogels
,”
RSC Adv.
5
,
106094
106097
(
2015
).
32.
X.
Li
,
S.
Chen
,
J.
Li
,
X.
Wang
, and
J.
Zhang
, “
3D culture of chondrocytes in gelatin hydrogels with different dtiffness
,”
Polymers
8
(
8
),
269
(
2016
).
33.
E.
Hoch
,
T.
Hirth
,
G. E. M.
Tovar
, and
K.
Borchers
, “
Chemical tailoring of gelatin to adjust its chemical and physical properties for functional bioprinting
,”
J. Mater. Chem. B
1
,
5675
5685
(
2013
).
34.
A.
Muench
,
C. P.
Please
, and
B.
Wagner
, “
Spin coating of an evaporating polymer solution
,”
Phys. Fluids
23
(
10
),
102101
(
2011
).
35.
V.
Garzo
and
M. L.
de Haro
, “
Kinetic models for diffusion in shear flow
,”
Phys. Fluids A
4
(
5
),
1057
1069
(
1992
).
36.
J.
Giacomin
,
R.
Bird
,
L.
Johnson
, and
A.
Mix
, “
Large-amplitude oscillatory shear flow from the corotational Maxwell model
,”
J. Non-Newtonian Fluid Mech.
166
,
1081
1099
(
2011
).
37.
L.
Martinetti
and
R. H.
Ewoldt
, “
Time-strain separability in medium-amplitude oscillatory shear
,”
Phys. Fluids
31
,
021213
(
2019
).
38.
C. H.
Lee
,
Y.
Lu
, and
A. Q.
Shen
, “
Evaporation induced self assembly and rheology change during sol-gel coating
,”
Phys. Fluids
18
,
052105
(
2006
).
39.
T.-H.
Kim
,
J.
Kim
, and
H.-Y.
Kim
, “
Evaporation-driven clustering of microscale pillars
,”
Phys. Fluids
28
(
2
),
022003
(
2016
).
40.
F.
Hidri
,
N.
Sghaier
,
H.
Eloukabi
,
M.
Prat
, and
S. B.
Nasrallah
, “
Porous medium coffee ring effect and other factors affecting the first crystallisation time of sodium chloride at the surface of a drying porous medium
,”
Phys. Fluids
25
,
127101
(
2013
).
41.
C. N.
Kaplan
,
N.
Wu
,
S.
Mandre
,
J.
Aizenberg
, and
L.
Mahadevan
, “
Dynamics of evaporative colloidal patterning
,”
Phys. Fluids
27
,
092105
(
2015
).
42.
J.
Fan
,
H.
Wu
, and
F.
Wang
, “
Evaporation-driven liquid flow through nanochannels
,”
Phys. Fluids
32
,
012001
(
2020
).
43.
Y.
Ichikawa
,
K.
Yamamoto
,
M.
Yamamoto
, and
M.
Motosuke
, “
Near-hydrophobic-surface flow measurement by micro-3D PTV for evaluation of drag reduction
,”
Phys. Fluids
29
,
092005
(
2017
).
44.
Y.
Liu
and
R.
Pandey
, “
Sol–gel phase transitions in thermoreversible gels: Onset of gelation and melting
,”
J. Chem. Phys.
105
(
2
),
825
836
(
1996
).
45.
H. B.
Bohidar
and
S. S.
Jena
, “
Kinetics of sol–gel transition in thermoreversible gelation of gelatin
,”
J. Chem. Phys.
98
(
11
),
8970
8977
(
1993
).
46.
D.
Levy
and
M.
Zayat
,
The Sol–Gel Handbook
(
Wiley VCH
,
Weinheim
,
2015
).
47.
H. E.
Park
,
J. M.
Dealy
,
G. R.
Marchand
,
J.
Wang
,
S.
Li
, and
R. A.
Register
, “
Rheology and structure of molten, olefin multiblock copolymers
,”
Macromolecules
43
,
6789
6799
(
2010
).
48.
A.
Giacomin
,
R.
Bird
,
C.
Aumnate
,
A.
Mertz
, and
A. M. A. M.
Schmalzer
, “
Viscous heating in large-amplitude oscillatory shear flow
,”
Phys. Fluids
24
(
10
),
103101
(
2012
).
49.
H. E.
Park
,
S. T.
Lim
,
H. M.
Laun
, and
J. M.
Dealy
, “
Measurement of pressure coefficient of melt viscosity: Drag flow versus capillary flow
,”
Rheol. Acta
47
,
1023
1038
(
2008
).
50.
M.
Nooranidoost
,
D.
Izbassarov
,
S.
Tasoglu
, and
M.
Muradoglu
, “
A computational study of droplet-based bioprinting: Effects of viscoelasticity
,”
Phys. Fluids
31
,
081901
(
2019
).
51.
J. T.
Cirulis
,
F. W.
Keeley
, and
D. F.
James
, “
Viscoelastic properties and gelation of an elastin-like polypeptide
,”
J. Rheol.
53
(
5
),
1215
1228
(
2009
).
52.
J. M.
Dealy
and
J.
Wang
,
Melt Rheology and its Applications in the Plastics Industry
, 2nd ed. (
Springer
,
Dordrecht
,
2019
), p.
80
.
53.
J. D.
Ferry
,
Viscoelastic Properties of Polymers
, 3rd ed. (
John Wiley & Sons
,
New York
,
1980
), p.
530
.
54.
C.
McIlroy
, “
A fundamental rule: Determining the importance of flow prior to polymer crystallization
,”
Phys. Fluids
31
,
113103
(
2019
).
55.
H. T.
Peng
and
P. N.
Shek
, “
Novel wound sealants: Biomaterials and applications
,”
Expert Rev. Med. Devices
7
(
5
),
639
659
(
2010
).
56.
E.
Biazar
,
M.
Najafi S.
,
S.
Heidari K.
,
M.
Yazdankhah
,
A.
Rafiei
, and
D.
Biazar
, “
3D bio-printing technology for body tissues and organs regeneration
,”
J. Med. Eng. Technol.
42
(
3)
,
187
202
(
2018
).
57.
M. J.
Lerman
,
J.
Lembong
,
G.
Gillen
, and
J. P.
Fisher
, “
3D printing in cell culture systems and medical applications
,”
Appl. Phys. Rev.
5
,
041109
(
2018
).
58.
L. S.
Nair
,
C. T.
Laurencin
, and
M.
Tandon
, “
Injectable hydrogels as biomaterials
,” in
Advanced Biomaterials; Fundamentals, Processing, and Applications
(
John Wiley & Sons
,
Hoboken
,
2000
).
59.
S.
Asliyuce
,
L.
Uzun
,
A. Y.
Rad
,
S.
Unal
,
R.
Say
, and
A.
Denizli
, “
Molecular imprinting based composite cryogel membranes for purification of anti-hepatitis B surface antibody by fast protein liquid chromatography
,”
J. Chromatogr. B
889-890
,
95
102
(
2012
).
60.
M.
Sirousazar
,
M.
Forough
,
K.
Farhadi
,
Y.
Shaabani
, and
R.
Molaei
, “
Hydrogels: Properties, preparation, characterization and biomedical applications in tissue engineering, drug delivery and wound care
,” in
Advanced Healthcare Materials
(
John Wiley & Sons
,
Hoboken
,
2014
).
61.
H.
Jianga
,
W.
Sua
,
P. T.
Mathera
, and
T.
Bunning
, “
Rheology of highly swollen chitosan/polyacrylate hydrogels
,”
Polymer
40
,
4593
4602
(
1999
).
62.
J. M.
Dealy
,
D. J.
Read
, and
R. D.
Larson
,
Structure and Rheology of Molten Polymers: From Structure to Flow Behavior and Back Again
, 2nd ed. (
Hanser Publications
,
Munich
,
2018
), p.
429
.