Gelatin Capsule Shell Cross-Linking

May 2, 2011
Pharmaceutical Technology
Volume 35, Issue 5

The authors develop a pratical approach to avoid unwanted interactions between pepsin and SLS in dissolution Tier II tests.

Hard gelatin capsules are a common solid oral dosage form, but exposure to accelerated conditions, e.g. 40 °C and 75% relative humidity (RH), can cause capsule shell cross-linking. Capsule shell cross-linking arises from gelatin polymerization, a process facilitated by high temperature, high humidity, ultraviolet (UV) and visible irradiation, dyes, and aldehydes (1–4). The main impact of gelatin cross-linking is prolonged capsule disintegration time, and a subsequent slow-down of drug product dissolution rate. In the event that hindered dissolution arises from gelatin cross-linking and the product fails specification, USP <711> recommends the addition of enzymes (e.g., pepsin) to the dissolution medium to serve as the Tier II dissolution test (5). It is important, though, to confirm that dissolution failure is a direct outcome of cross-linked gelatin shells rather than degradation of drug product performance.

Sodium lauryl sulfate (SLS) is a surfactant commonly used in dissolution medium to improve the solubility of poorly water-soluble drugs. The presence of SLS in dissolution medium deactivates pepsin, which complicates the Tier II method described above (6). One option would be to redevelop the dissolution method and abandon SLS. But this option could be costly in time and resources, and may discourage the use of SLS in capsule formulations in general, despite its excellent solubilizing capability, low cost and ease of use. Performing Tier II dissolution tests in the presence of SLS is considered beneficial to the development and quality control of capsule formulations.

This article will detail the experimental procedures and the study results of a case where the above issues were encountered and tackled in the development of a capsule formulation. A slowdown in dissolution rate was discovered for the gelatin capsule formulation when it was stored at accelerated conditions of 40 °C and 75% RH for three months.


Size one opaque hard gelatin capsule shells were purchased from Capsugel. Dissolution was performed using USP Apparatus II (paddles), Model VK 7000 (Varian). Stand-alone UV–Vis spectrometer with diode array capacity, Model 8453, was from Agilent. Capsule sinkers (size 8/23) were from Sotax. SLS (reagent grade > 99%) was purchased from Fisher Scientific. Full flow cannula filters (10 µm) were from Quality Lab Accessories. Pepsin (800–2,500 units/mg) purified from porcine gastric mucosa was purchased from Sigma-Aldrich. All other chemicals were ACS grade or equivalent.


Dissolution methods. Tier I dissolution was performed using USP paddle apparatus in 900 mL of 0.01 N HCl with 1.0% SLS in each vessel at 37 °C. The paddle rotation speed was 75 rpm.

Samples were obtained at predetermined time points of 10, 20, 30, 45, and 60 min. After 60 min the paddle speed was increased to 250 rpm for another 15 min before samples at the "infinity" time point were withdrawn. All samples were analyzed using a UV–Vis spectrometer at a wavelength of 266 nm.

The initial Tier II dissolution method was developed following USP <711>, using a premixed medium containing 900 mL of 0.01 N HCl, 1.0% SLS and 750,000 units/L purified pepsin in each vessel.

The final Tier II dissolution method was modified from USP <711> by using 600 mL of 0.01 N HCl solution containing 750,000 units/L of purified pepsin in each vessel at the beginning of the test. After 5 min, an additional 300 mL of 0.01N HCl solution containing 3.0% SLS was added to each vessel. This medium was preheated and kept at 37 °C before transferring. Other method conditions were constant

Capsule switching test procedure. To identify the cause of dissolution slowdown, a capsule switching test was conducted. The contents of six capsules, which had been stored at 40 °C and 75% RH for three months and showed dissolution slow down, were fully transferred into six fresh shells. The fresh capsule shells were from the same batch as those used in the stability study and were stored in a closed container at ambient conditions. The six emptied (i.e., aged) capsule shells, on the other hand, were refilled with a fresh drug blend made with the same formula and manufacturing process.

Table I: Dissolution results for capsules stored at 40 °C and 75% relative humidity (RH) for 1 and 3 months.

Results and discussion

Table I and Figure 1 provide the dissolution results and profiles of capsules stored at 40 °C and 75% RH for one and three months, respectively. Testing was performed using the Tier I method. Results should conform to a Q value of 70% at 45 min. Comparing the two sets of data, it is clear that capsules stored for three months had significant variation. Four out of twelve capsules had release of 73.0%, 66.2%, 47.7%, and 53.1%, respectively. The results did not meet Stage I or Stage II criteria.

Figure 1: Dissolution profiles for capsules stored at 40 °C and 75% relative humidity for 1 and 3 months.

The observation of reluctant capsule shell rupture was a good indication that dissolution failure was most likely caused by cross-linked capsule shell rather than drug performance. To further confirm this theory, an investigation in which capsules were switched and subjected to dissolution testing using Tier I was performed. Dissolution data for switched capsules are provided in Table II; data for fresh drug blend in fresh capsule shells are included for comparison. As expected, the capsules with fresh drug blend in aged capsule shells had individual low results and significantly high variation at every time point. The capsules with either old or fresh blend in fresh capsules shells both had satisfactory results. The study results proved that the aged capsule shells, rather than product-quality change, caused the original dissolution failure.

Table II: Dissolution result comparison of different capsule samples.

For hard gelatin capsules that do not conform to dissolution specification, USP <711> suggests that the test is repeated with the addition of purified pepsin that results in an activity of 750,000 units or less per 1000 mL to the medium that has a pH of less than 6.8 (5). Therefore, another six capsules from the original three-month 40 °C and 75% RH storage were tested using the initial Tier II method with pre-mixed medium containing 900 mL of 0.01 N HCl, 1.0% SLS and 750,000 units/L purified pepsin in each vessel. The medium was freshly prepared. The results are provided in Table III. On visual observation, the capsule disintegrated slowly. Some capsules appeared to be gelling with blend trapped inside during the test until a high paddle speed of 250 rpm at "infinity" mechanically ruptured them. The dissolution was slow; the results did not conform to a Q value of 70% at 45 min and displayed high standard deviations. In this case, the presence of SLS may have deactivated pepsin as reported.

Table III: Dissolution results of coaddition of pepsin and SLS in the medium.

To remove the effect of SLS on capsule shell disintegration, Medium #1 was prepared consisting of 0.01 N HCL with 750,000 units/L pepsin without the addition of SLS. Tier II dissolution was performed with 600 mL of Medium #1/vessel. Two minutes into the run, all six capsules were observed to be fully disintegrated. At 5 min, 300 mL of prewarmed Medium # 2, consisting of 0.01 N HCL with 3% SLS, was transferred into each running vessel without disturbing the dissolution run. The final composition of the resulting total medium was 0.01 N HCL with 1% SLS and 500,000 units/L pepsin. The dissolution results and profiles are provided in Table IV and Figure 2, respectively. Satisfactory results were obtained, with tight standard deviations.

Table IV: Dissolution results of stepwise addition of pepsin and SLS in the medium.

The results indicated that stepwise addition of pepsin and SLS enabled both agents to take effect individually and sequentially in the dissolution medium. Pepsin digested the cross-linked capsule shells at the beginning, whereas the addition of SLS afterwards increased drug solubility and wettability. Therefore, the addition of SLS to the dissolution medium need not be discouraged when developing dissolution methods for capsule formulations. SLS is commonly included as a wetting agent inside the capsule formulation; this practice should not be affected by the results of this study, because SLS deactivation of pepsin was observed outside of the capsule in the dissolution medium before dissolution took place. By taking a stepwise addition approach, once the cross-linked capsule shell ruptures and dissolution starts, SLS inside the formulation will work as expected.

Figure 2: Dissolution profile comparison of coaddition and stepwise addition of pepsin and sodium lauryl sulfate (SLS).

The 5-min time delay between the addition of pepsin and SLS was further confirmed to be sufficient using more severely stressed capsules. The Tier II method was fully validated for linearity, specificity, accuracy, repeatability, intermediate precision, and stability of standard and sample solutions.


Gelatin capsule shell cross-linking is a common problem for a capsule formulation during stress or stability studies at accelerated storage conditions. Switching the stressed capsule shells and blends with fresh ones can easily prove that the shells are the cause of slowed dissolution. SLS deactivates pepsin despite its advantages and wide use as a surfactant. However, stepwise addition of pepsin and SLS respectively enables each agent to take effect separately. Therefore, SLS need not be abandoned during dissolution method development for gelatin capsule formulations.


The authors are grateful to Chris Connolly, manager of analytical development; Jack Chen, supervisor of analytical development; and Andy Cleaver, supervisor of analytical development, for coordinating tasks. We are also grateful to Wendy Liu, chemist in analytical development, for conducting dissolution tests, and Rajeev Bhatnagar, senior technical project lead, for providing fresh capsule blends.

Xiling Song* is senior research associate, Yong Cui is scientist, and Minli Xie is senior scientist, all at Small Molecule Pharmaceutical Sciences, Genentech, 1 DNA Way, South San Francisco, CA 94080, *To whom all correspondence should be addressed. Submitted: Nov. 10, 2010. Accepted: Feb. 7, 2011


1. S. Singh, R. Manikandan, and S. Singh, Pharm. Technol. 24 (5), 58–72, (May 2000).

2. K.S. Murthy, N.A. Enders, and M.B. Fawzi, Pharm. Technol. 13 (3), 72–86 (1989).

3. K.S. Murthy, R.G. reisch, Jr., and M.B. Fawzi, Pharm. Technol, 13 (6), 53–58 (1989).

4. S. Singh et al., Pharm. Technol. 26 (4) 36–58, (2002).

5. USP33–NF28 (US Pharmacopeial Convention, Rockville, MD, 2010), p. 719.

6. M.F.S. Whisler, J. S. Staton, and R. B. DePrince, poster presentation at AAPS conference (New Orleans, LA, 1999).

Citation: When referring to this article, please cite it as "X.Song, Y.Cui, M.Xie, "Gelatin Capsule Shell Cross Linking," Pharmaceutical Technology 35 (5) 62–68 (2011)."