Filling of High-Concentration Monoclonal Antibody Formulations into Pre-filled Syringes: Investigating Formulation-Nozzle Interactions To Minimize Nozzle Clogging

Suck-Back (SB) Pattern of Tapered Polypropylene versus Straight Glass and Stainless Steel Nozzles

SB experiments were initially performed on the mAb formulation (200 mg/mL with a viscosity of 9 cP) using three 2.0-mm nozzles: a stainless steel nozzle, a glass nozzle, and a polypropylene nozzle (Figure 1). The polypropylene nozzle has a gradual tapering shape, while the shape of the glass and the stainless steel nozzles is straight. A 180 min drying experiment was performed on the mAb formulation in all three nozzles at different SB settings (SB0 to SB3). Liquid was dispensed at 15, 30, 60, and 180 min to observe if the nozzle was clogged. The results shown in Table I (results for the Teflon-coated stainless steel nozzle the and silicone-coated glass nozzle will be discussed later) suggested that the glass nozzle is most likely to be clogged in 15 or 30 min at all SB settings, while the polypropylene nozzle is least likely to clog (no clogging after 180 min). The stainless steel nozzle is also easy to clog within 30 min at all SB settings except for SB2 (clogged at 60 min). The presence of an optimal SB setting (i.e., a narrow SB range) for stainless steel and glass nozzles is consistent with previous findings (1).

The SB pattern of the mAb formulation in the glass and the plastic nozzles was visualized with a camera. Interestingly, different SB patterns were observed: a liquid plug accumulated at the tip of the glass nozzle versus a liquid plug floating between the nozzle tip and the meniscus of the sucked-back liquid in the plastic nozzle (Figure 2). The delay of formulation drying and subsequent nozzle clogging in the polypropylene nozzle may be attributed to this unique SB pattern where the liquid plug formed away from the nozzle tip. Inside the nozzle, air flow becomes stagnant and the rate of water evaporation slows down. Based on these initial results, two likely root causes for the unique SB pattern were identified: (1) the tapering shape and (2) the interfacial properties between the liquid and the nozzle material.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Suck-back (SB) pattern of the mAb formulation in polypropylene and glass nozzles at different settings: (a) SB2 and (b) SB4.

Investigating the Effect of the Shape Factor Using a Glass Pipette

In order to evaluate whether or not filling nozzle shape alone could result in the unique SB pattern observed with the tapered polypropylene nozzles, the SB pattern of the mAb formulation at different SB settings was evaluated in glass pipettes with sharp, tapering tips (Figure 3a for 1 mm ID tip and Figure 3b for 2 mm ID tip). The data suggested that the air bubble created by SB at the tip tended to rise up, which was particularly true in the cases of narrow tip opening (1 mm ID) and large air bubble (high SB settings). At the tip region (sharp tapering) the shape of the air bubble was not uniform (deformed), so the bubble moved to a zone where the pressure is lower to form a more uniform bubble shape. As the bubble moved up, the pipette tip was filled with liquid. This observation indicated that a nozzle with a tapering tip may actually discourage a “clean” SB pattern (i.e., the absence of a liquid plug at the nozzle tip), and that shape was likely not a large contributor toward the SB pattern observed in the tapered polypropylene nozzles.

• Download figure
• Open in new tab
• Download powerpoint

Figure 3

Suck-back (SB) pattern of the mAb formulation in a glass pipette with (a) 1 mm ID tip and (b) 2 mm ID tip at various SB settings.

Investigating the Effect of Nozzle Material Hydrophobicity on Suck-back (SB) Behaviors

In order to evaluate the effect of the interfacial relationship between the filling nozzle material and formulation, SB behavior in different substrate materials of differing hydrophobicity were evaluated. Hydrophilicity and hydrophobicity are often termed to interpret the interactions between a liquid and its substrate (nozzle material). The substrate surface is hydrophilic if a hydrophilic liquid (e.g., water) tends to spread over it, whereas with a hydrophobic surface a hydrophilic liquid tends to reduce its surface area of contact. Contact angle can conveniently quantify hydrophilicity and hydrophobicity of the substrate. This study assessed five substrate materials: glass, stainless steel, three plastics (polypropylene, silicone, and Teflon), and three liquids (purified water, a mAb formulation, and a 60% sucrose solution) to understand the impact of nozzle material hydrophobicity on SB behavior. Table II summarizes the contact angle measurements and liquid viscosity data for the materials and liquids assessed in this investigation. The increasing contact angle of water, going from on glass (<20°), stainless steel (68°), and plastics (97–122°) suggested that the glass surface is the most hydrophilic and that all three plastics are hydrophobic. Among the three plastic materials, Teflon appears to be most hydrophobic. The contact angle of the mAb formulation was lower than that of water on all substrates, due perhaps to the presence of surfactant (polysorbate 20) and surface properties of the protein, while the contact angle of the 60% sucrose solution was comparable to that of water.

To verify the effect of surface hydrophobicity, the SB behaviors of all three liquids in a glass nozzle (3.0 mm ID) and a piece of Teflon tubing (3.2 mm ID) were visually compared at various SB settings. Figure 4 shows the photographs highlighting SB performance at a low SB setting (SB3) and a high SB setting (SB9). For water (Figure 4a), no liquid plug was observed in either nozzle, except for a slight accumulation of water at the tip of the glass nozzle at the high SB setting. For the more viscous mAb formulation (Figure 4b), neither nozzle showed a liquid plug at SB3. Although a liquid plug was formed at SB9 in both nozzles, it is located at the nozzle tip for the glass nozzle and within the tip for the Teflon nozzle. To suck back the most viscous sucrose solution (Figure 4c), SB3 could barely withdraw the liquid into the glass nozzle and only slightly into the Teflon nozzle. At SB9, a liquid plug formed in both nozzles with the same pattern as previously observed—at the tip of the glass nozzle and away from the tip of the Teflon nozzle. The SB behavior for the three different liquids at SB9 for both nozzles was repeatedly tested and found to be reproducible. The results confirmed that at high SB, a liquid plug tends to form at the tip of a hydrophilic nozzle but stays away from the tip of a hydrophobic nozzle.

• Download figure
• Open in new tab
• Download powerpoint

Figure 4

Suck-back/meniscus pattern of (a) water, (b) mAb formulation, and (c) 60% sucrose solution in glass and Teflon nozzles at SB3 and SB9.

Another observation also confirmed the hydrophobic nature of the Teflon nozzle: its convex meniscus versus the concave meniscus of the glass nozzle (indicated in Figure 4). A convex meniscus occurs when liquid molecules have a stronger attraction to each other (cohesion) than to the material of the container (adhesion) (7), that is, liquid retreating from the Teflon surface. A liquid whose molecules have a stronger attraction to the material of the container than to each other forms a concave meniscus, that is, the liquid climbing up the glass surface.

The relationship of SB volume and SB setting for all three liquids is presented in Figure 5. The trend of higher SB volume in response to lower viscosity was consistent with previous findings (1). The Teflon nozzle resulted in a slightly larger SB volume than the glass nozzle, suggesting that the hydrophobic surface promoted liquid SB.

• Download figure
• Open in new tab
• Download powerpoint

Figure 5

Suck-back volume of water, mAb formulation, and 60% sucrose solution in glass and Teflon nozzles in response to SB setting (SB0, SB3, SB6, and SB9).

Effect of Liquid Viscosity and Nozzle Material Hydrophobicity on Liquid Plug Formation

The formation of a liquid plug in the filling nozzle depended on the thickness of the liquid film on the inner surface of the nozzle. The liquid film thickness depended on liquid viscosity; the more viscous liquids left a thicker film. For example, in the case of water (Figure 4a) at the high SB setting (the worst case), the amount of water accumulated at the tip of the glass nozzle was too small to form a liquid plug. However, for the more viscous mAb formulation and the 60% sucrose solution, a liquid plug formed and its volume increased as the liquids were sucked back.

The formation of a liquid plug at different locations in the nozzle can be explained by capillary action, which represents a force balance between the cohesive force of the liquid (surface tension), the adhesive force between the liquid and the nozzle material, and the force of gravity. The liquid film left on the internal surface of a hydrophilic nozzle initially remained intact upon SB. However, the film began to flow down due to gravity and stopped at the nozzle tip due to surface tension and adhesive force. In a hydrophobic nozzle, the film quickly contracted toward between the nozzle tip and the meniscus of the sucked-back liquid because the liquid will resist association with the hydrophobic surface. The final location of the liquid plug was balanced by the force of gravity and the uplifting repulsive force of the hydrophobic surface below the liquid plug.

Relationship of Nozzle Material Hydrophobicity and Nozzle Clogging

Based on the observations of the effect of nozzle material hydrophobicity on SB behavior, it was hypothesized that the material of the hydrophobic nozzle tip plays a dominant role in slowing down formulation drying and nozzle clogging due to surface interactions between the liquid and the nozzle material. This hypothesis was further confirmed by expanding the 180 min drying experiment to include two new hydrophobic nozzles (Teflon-coated stainless steel [Figure 1a] and siliconized glass) (Table I). As predicted, the behavior of the drug product formulation in the Teflon-coated stainless steel and siliconized glass nozzles mimicked the behavior of the previously tested polypropylene nozzle. No clogging was observed for most SBs after drying for 3 h, compared with the hydrophilic nozzles (stainless steel and glass), which were completely clogged within 30 min.

The extended drying time can be largely attributed to the differences in SB behavior noted earlier. However, even when a liquid droplet was formed at the tip of a hydrophobic nozzle and dried after an extended interruption, no clogging was observed (Table I). This observation can also likely be attributed to the nozzle material's hydrophobicity. A liquid plug at the tip of a hydrophobic nozzle is less likely to completely clog the nozzle since the liquid plug is less likely to firmly adhere to a hydrophobic nozzle surface compared to a hydrophilic nozzle.

Relationship of Formulation Composition and Nozzle Clogging

For nozzle clogging to occur, the front edge of the formulation in the nozzle has to be dried, hardened, and firmly attached to the nozzle surface. Nozzle clogging is unlikely if any one of the three events does not occur. Dehydration (water evaporation) of the formulation front edge should take place during filling process interruption, typically less than 1 h. In addition to environmental conditions (temperature, humidity, air flow, etc.) and nozzle material hydrophobicity, the drying rate of a liquid can be affected by the composition of the formulation as well.

When a drop of the mAb formulation (200 mg/mL) and the 60% sucrose solution were left to dry on a stainless steel plate for up to 24 h, the mAb formulation dried faster (in 3 h) than the sucrose solution (in 24 h) (Figure 6). Other than a slower drying rate, during the history of drying, the surface of the drop of the sucrose solution turned viscous and sticky, but it remained in the liquid form. On the contrary, the surface of the drop of the mAb formulation turned to a dried crust quickly. It has been reported that upon spray drying, the droplets of sugar-rich solutions are sticky and tend to stick to the dryer wall or collide with each other to form undesired agglomerates (8, 9), whereas the droplets of mAb formulations are not sticky in the drying chamber of the spray dryer and can be effectively collected (10). However, as expected, the stickiness of the mAb droplets increased after an increasing amount of sugar was added to the mAb formulation (10). It can be hypothesized that the mAb formulation easily forms a dry crust at the air–liquid interface, which gradually thickens and turns into a hardened mass upon further water evaporation. The nozzle will not clog if the hardened mass can be dislodged under the pressure of pump dispensing, or in other words, the dry material is not firmly attached to the nozzle material.

• Download figure
• Open in new tab
• Download powerpoint

Figure 6

Drying behavior of a liquid drop of (a) 60% sucrose solution and (b) mAb formulation on a stainless steel sheet as a function of time (T = 0, 3, and 24 h).