• inverse gas chromatography;
  • surface free energy parameters;
  • pharmaceutical powders;
  • HPMC;
  • microcrystalline cellulose;
  • magnesium stearate


  1. Top of page
  2. Abstract

The use of inverse gas chromatography to assess surface properties of a range of pharmaceutical powders was examined. The powders were two sources of hydroxy propylmethyl cellulose (HPMC), microcrystalline cellulose, magnesium stearate, and acyclovir. These were selected to cover a range for properties from amorphous to crystalline, hydrophilic to hydrophobic, and high to low aqueous solubility. It was found that many powders gave a similar value for the dispersive surface energy, which is surprising given the differences in chemical nature. It is likely that this is due to the use of infinite dilution giving rise to the study of specific regions of the powder surface only. The values obtained for dispersive energies were not influenced by packing mass or flow rate of the carrier gas. The retention of polar probes on the column was a concern for the amorphous HPMC samples. This gave rise to derived values for acid-base nature which varied depending on sample mass and carrier gas flow rate. The data show that care must be taken when studying amorphous samples for which it is possible to obtain diffusion into the material rather than just surface adsorption of probes. Despite these problems, it was still possible to differentiate between the samples (including differences between the two HPMC samples) by use of polar probes. It was also possible to see differences in absorption into the sample, reflecting the different physical forms. For example, microcrystalline cellulose behaved very differently to HPMC. It can be concluded that inverse gas chromatography is a valuable characterization tool, but it must be used with care especially with respect to polar probes on amorphous samples. © 2003 Wiley-Liss, Inc. and the American Pharmaceutical Association J Pharm Sci 92:1286–1294, 2003


  1. Top of page
  2. Abstract

The surface properties of powders will influence the way in which components of formulations interact, and consequently are important in the manufacture and use of medicines. It is therefore important to know the nature and magnitude of solid surface free energy. Several methods for surface energy determination are available and have been reviewed by Buckton1; however, all of the methods give rise to certain problems for powdered systems.

In recent years, there has been growing interest in the use of inverse gas chromatography (IGC) for the assessment of the surface energy of powders. IGC is appealing because it is well suited to powdered samples and the basic concept is that of gas chromatography (GC), for which there is good understanding. The only difference between IGC measurements and conventional GC is that with IGC the unknown surface is the powder that is packed into a column and the known materials are the vapor probes that are injected (i.e., the inverse of the known and unknown in conventional GC). It is understandable that the clear need to study powder surfaces and the promise of a method based on a tried and tested concept such as GC, will give rise to a great interest in, and rapid take up of, IGC. It is no surprise then that the number of publications in this field is rising rapidly, including those which have shown potential for IGC use in batch to batch variability,2 the influence of milling on the surface free energy,3,4 differences of surface free energy for two isomers,5 and the influence of humidity on surface free energy of different powders.6,7

As IGC becomes more widely used, it is essential that there is a clear understanding of the advantages and possible limitations of the method. The aim of the current study was to use materials that are common pharmaceutical excipients that cover a range of properties, i.e., two forms of hydroxypropyl methyl cellulose (HPMC), microcrystalline cellulose, magnesium stearate, and a model hydrophilic drug substance. These samples were selected because HPMC is amorphous, microcrystalline cellulose is part crystalline, highly porous particle, and contains residues of amorphous content, magnesium stearate is hydrophobic and predominantly crystalline (although not pure), and the model drug in contrast is water soluble and crystalline. This series of samples gives scope to understand any difficulties that can be seen when dealing with amorphous/complex and crystalline samples that range from hydrophilic to hydrophobic in nature.


  1. Top of page
  2. Abstract

The following powders were investigated: Pharmacoat 606 (HPMC, Syntapharn, Germany), Methocel K4M (HPMC, Colorcon, UK), Avicel PH101 (microcrystalline cellulose, FMC), magnesium stearate (Aldrich, Germany), and acyclovir (Lek, Slovenia). All the powdered samples were sieved and the fraction of particle size from 63 to 125 μm was used for the analysis [although micrographs (not shown) revealed that the magnesium stearate was very fine and existed as aggregates of this sieve size, whereas the other materials were individual particles of this size]. Vapors for retention time analysis and calculation of surface energy parameters were methane (research grade, BOC), hexane (99%+, Sigma-Aldrich), heptane (high-performance liquid chromatography grade, Fisher), octane (99%+, Lancaster), nonane (99%+, Sigma-Aldrich), decane (99%+, Acros), ethanol, ethyl acetate (99%+, Sigma-Aldrich), chloroform (99%, BDH), and acetone (99%+, Acros).

IGC Measurements

A commercial inverse gas chromatograph (Surface Measurements Systems Ltd.) was used for powder analysis. A Hewlett Packard 6890 series gas chromatograph oven was used to control the probe liquids at a temperature of 28°C. The 6890 GC data acquisition system was used to record data from the flame ionization detector at the column outlet. A separate, purpose built column oven was used to control the sample (column) temperature at 25°C. Two glass silanized (dymethyldichlorosilane; Repelcote BDH, UK) columns (inside diameter = 3 mm) were packed with a known mass of the powder sample by vertical tapping for 10 min. Dymethyldichlorosilane silanized glass wool was put on both sides of the column to support the powder samples. Helium at different flow rates (7, 11, 15, 25, and 35 mL/min) was used as a carrier gas. Packed columns were conditioned for 5 h at 25°C before measurement (with the flow of dry carrier gas). The IGC system was fully automated with purpose written control software (SMS iGC Controller v1.3) and data were analyzed using SMS iGC Analysis macros. Infinite dilution was demonstrated by injecting increasing concentrations of probe gas and ensuring that the retention behavior remained unchanged.

Elution Efficiency

Vapor probes were injected at a concentration of 2% v/v (using head space sampling) into an empty glass column and into columns containing the test powders; peak areas were computed. The elution efficiency of each probe was calculated as the ratio of the peak areas for the packed and empty columns (minimum n = 3).

Specific Surface Areas

The specific surface areas of the samples were calculated using multipoint Brunauer-Emmet-Teller nitrogen adsorption at 77 K. The TriStar 3000 gas adsorption analyzer (Micromeritics Instrument Corp., Norcross, GA) was used for the measurement.

Surface Energy Calculations

According to Fowkes,8 the surface free energy (γs) of solid consists of two components: the dispersive component (γmath image) describing London type of interactions, and the specific component γmath image including all other types of interaction (H-bonding, acid-base, polar). Similarly, the specific energy of adsorption can be split into dispersion and specific components:

  • equation image(1)
Dispersive Surface Free Energy

The method used was as proposed by Saint and Papirer.9The dispersive surface energy of the solid is obtained from:

  • equation image(2)

With the retention volume of n-alkanes being obtained from:

  • equation image(3)

where Vn = retention volume of the injected probe; j = correction factor due to pressure differences of the carrier gas on both sides of the column; F = flow rate of the carrier gas; tr = retention time of the injected probe; and t0 = retention time of the noninteracting standard.

j is calculated by use of:

  • equation image(4)

where Pin = pressure at the inlet of the column and Pout = pressure at the outlet of the column.

From a plot of the free energy of adsorption of liquid probes onto solids (RTlnVn) as a function of the γ a straight line is obtained. The dispersive component of the solid surface energy is then obtained from the slope of this line.

Acid and Basic Parameters of Solid Surface Energy

On the plot of RTlnVn as a function of the polar probes will be displaced from the linear response seen for the alkanes. The vertical difference between the polar probe and alkane line gives the specific energy of interaction ΔGmath image. The surface acceptor (Ka) and donor (Kd) natures are related to ΔGmath image by the following equation10:

  • equation image(5)

where DN and AN* are liquid donor and acceptor numbers of probes.

The coefficients from eq. 5 can be determined by the least square method from the linear relationship for the series of polar probes characterized by different DN and AN* values (eq. 6) determined by Gutmann11 and Riddle and Fowkes.12

  • equation image(6)


  1. Top of page
  2. Abstract

What is the Meaning of Dispersive Surface Energy Measured at Infinite Dilution?

Alkane probes were used to determination the dispersive surface energy of the powders (n = 8). Retention peaks for the alkanes were symmetrical and reproducible (not shown). The dispersive surface energy for Methocel (mean 52 mJ/m2, Table 1) was greater than for Pharmacoat (mean 47 mJ/m2, Table 2), which demonstrates that on occasions it may be possible to differentiate between similar materials by use of dispersive surface energy. However, the other powders analyzed had surprisingly similar values of dispersive surface energy, i.e., 50 mJ/m2 for Avicel (Table 3) and magnesium stearate (Table 4) and 46 mJ/m2 for acyclovir (Table 5) are much more similar than may be expected given that Avicel PH101, magnesium stearate, and acyclovir are very different materials because they are respectively hydrophilic (but insoluble in water), hydrophobic, and reasonably hydrophilic with good water solubility. Given the nature of these materials, it may have been expected that the dispersive surface energies would be significantly different for each material. It is possible, therefore, that relatively similar values of the nonpolar parameter of surface energy will be obtained for most pharmaceutical powders because London interactions for substances that consist mostly of C, H, and O atoms can be similar for very different substances. A comparison of these results with literature values show other examples where the dispersive surface energy is around the same value: manitol βD 47.9 mJ/m2,5 salbutamol sulphate (unmilled) 49.7 mJ/m2,4 theophylline 50.8 mJ/m2,13 and zamifenacin 46.8 mJ/m2.14 However, very different results are also quoted: manitol DL 73.7 mJ/m2,5 salbutamol sulphate 58.6 (unmilled), 62.2 and 64.5 mJ/m2 (milled),4 caffeine 39.9 mJ/m2,13 starch 39.8 mJ/m2,14 and lactose monohydrate 31.3 mJ/m2.7 There is clearly a need to obtain a better understanding of exactly what is being measured when using IGC at infinite dilution. It certainly seems that very often a surface energy of around 50 mJ/m2 can be seen, but this is obviously not always the case. Infinite dilution experiments will probe a small (probably highest energy) region of the surface only. It is probable that this high energy region of the surface could dominate interfacial interactions involving the powder; however, if we are to understand the surface nature of materials properly, we will need more information than just the dispersive energy at infinite dilution.

Table 1. Surface Energy Parameters for Methocel K4M
Mass (mg)Flow (mL/min)γmath image (mJ/m2)ΔHAB (kJ/mol) AcetoneΔHAB (kJ/mol) ChloroformΔHAB (kJ/mol) Ethylacetate
250752.9 (1.1)19.6 (0.5)6.2 (0.4)19.0 (0.6)
2501151.1 (0.6)17.3 (0.2)4.8 (0.3)16.3 (0.2)
5001153.5 (0.7)20.9 (0.2)8.3 (0.3)20.5 (0.2)
5001553.4 (0.8)20.0 (0.4)7.2 (0.1)19.3 (0.1)
5002551.1 (1.3)18.2 (0.2)5.8 (0.1)17.4 (0.3)
5003550.2 (0.7)17.1 (0.2)5.1 (0.1)16.1 (0.3)
Table 2. Surface Energy Parameters for Pharmacoat PH606
Mass (mg)Flow (mL/min)γmath image (mJ/m2)ΔHAB (kJ/mol) AcetoneΔHAB (kJ/mol) ChloroformΔHAB (kJ/mol) Ethylacetate
  • a

    Particles smaller than 50 μm.

250746.0 (0.5)19.9 (0.2)10.9 (0.4)20.0 (0.3)
2501146.1 (0.7)18.9 (0.3)9.1 (0.3)18.6 (0.4)
250a1148.3 (1.0)18.7 (0.5)9.6 (0.3)18.9 (0.4)
2501546.7 (0.6)17.9 (0.3)8.1 (0.3)17.6 (0.4)
5001147.2 (0.2)21.6 (0.4)13.4 (0.1)21.6 (0.3)
5001550.5 (0.1)21.7 (0.1)12.5 (0.1)21.8 (0.1)
5002548.1 (1.2)20.1 (0.4)10.9 (0.5)19.9 (0.5)
5003546.2 (1.3)19.0 (0.1)9.6 (0.2)18.4 (0.2)
Table 3. Surface Energy Parameters for Avicel PH101
Mass (mg)Flow (mL/min)γmath image (mJ/m2)ΔHAB (kJ/ mol) AcetoneΔHAB (kJ/ mol) EthanolΔHAB (kJ/mol) ChloroformΔHAB (kJ/mol) Ethylacetate
  • a

    Particles smaller than 50 μm.

250750.0 (0.4)13.4 (0.2)8.5 (0.3)0.1 (0.1)12.0 (0.1)
2501151.0 (0.6)13.3 (0.4)8.6 (0.1)0.1 (0.3)11.8 (0.1)
250a1149.3 (0.9)13.3 (0.4)9.1 (0.3)1.1 (0.2)11.9 (0.2)
2501550.9 (0.8)13.4 (0.1)8.3 (0.1)0.4 (0.1)11.7 (0.2)
5001549.3 (0.3)13.7 (0.1)9.7 (0.1)1.2 (0.0)12.5 (0.1)
5002549.4 (0.6)13.5 (0.2)9.7 (0.2)1.2 (0.1)12.4 (0.2)
5003549.3 (1.1)13.7 (0.2)9.7 (0.1)1.0 (0.2)12.6 (0.2)
Table 4. Surface Energy Parameters for Magnesium Stearate
Mass (mg)Flow (mL/min)γmath image (mJ/m2)ΔHAB (kJ/mol) AcetoneΔHAB (kJ/mol) ChloroformΔHAB (kJ/mol) Ethylacetate
  • a

    Particles smaller than 50 μm.

75749.9 (1.1)4.4 (0.3)0.9 (0.1)6.7 (0.1)
751149.8 (0.5)4.2 (0.1)0.8 (0.1)6.6 (0.1)
75a1149.8 (0.5)4.2 (0.1)0.8 (0.0)6.6 (0.1)
751550.1 (0.3)4.2 (0.1)0.9 (0.2)6.7 (0.1)
150747.8 (1.0)3.9 (0.2)0.7 (0.1)6.5 (0.1)
1501150.1 (0.2)4.0 (0.1)0.7 (0.1)6.5 (0.1)
1501550.3 (0.8)3.7 (0.2)0.7 (0.1)6.4 (0.1)
Table 5. Surface Energy Parameters for Acyclovir
Mass (mg)Flow (mL/min)γmath image (mJ/m2)ΔHAB (kJ/ mol) AcetoneΔHAB (kJ/ mol) EthanolΔHAB (kJ/mol) ChloroformΔHAB (kJ/mol) Ethylacetate
150745.3 (0.9)6.9 (0.3)9.7 (0.1)1.7 (0.1)9.1 (0.1)
1501145.9 (0.1)6.9 (0.1)9.9 (0.0)1.7 (0.1)9.3 (0.1)
1501547.4 (0.7)7.0 (0.1)10.2 (0.0)1.7 (0.2)9.7 (0.1)
250746.4 (1.4)7.3 (0.2)1.7 (0.1)9.4 (0.1)

The Influence of Powder Mass and Carrier Gas Flow Rate

Literature data show that flow rates used for IGC measurements can be very different (from 3 to 40 mL/min.15 Newell et al.7 have shown that the flow rate does not affect the surface energy when using crystalline lactose particles in IGC. However, this does not mean that flow rate will not affect the results for other samples. For example, with amorphous samples (such as celluloses), it is possible that probes can diffuse into the powder and therefore it is interesting to study the influence of flow rate.

HPMC and microcrystalline were analyzed with sample loads of 250 and 500 mg at flow rates from 7 to 35 mL/min (Tables 1 and 2). Magnesium stearate was analyzed by use of sample masses of 75 and 150 mg and acyclovir of 150 and 250 mg, because the pressure decrease with these two samples was too high when more powder was used. The pressure decrease across the column was different for different samples and was reasonably well correlated with specific surface areas of the samples (Table 6). This demonstrates that fine particles (such as the magnesium stearate) require lower mass to be loaded into the column.

Table 6. Correlation between Pressure Decrease and Specific Surface Area of the Analyzed Powdersa
SampleMass (mg)Pressure Decrease (Torr)Surface Area (m2/g)
  • a

    Flow rate in all cases 11 mL/min.

Pharmacoat 606250750.87 ± 0.02
Methocel K4M250441.14 ± 0.03
Avicel PH101250801.31 ± 0.01
Aciclovir1503352.06 ± 0.02
Magnesium stearate1505326.22 ± 0.07

The influence of different carrier gas flow rates and different powder masses can also be seen by consideration of Tables 1–5. According to literature, the IGC measurement error of dispersive surface energy can be as high as 2–3 mN/m.5,12,16 This means that nonpolar parameter differences measured in the current study can be attributed to statistical error. The data in Tables 2–5 and 7 show such variation in calculated dispersive surface energy and there is no obvious change in the value as a consequence of flow rate or powder load. It is reassuring, therefore, that even if it is unclear why many materials exhibit similar values of dispersive surface energy when using IGC at infinite dilution, the data are not affected by changes in powder mass or flow rate.

Table 7. The Percentage of Probe Molecules Eluted through the Column as Compared with the Amount of Probes Injected: Methocel Sample
Mass (mg)Flow (mL/min)Yield

Polar Probes

Whereas retention peaks for nonpolar probes were Gaussian for all the powders, nonsymmetrical peaks of polar probes on both samples of HPMC were observed (Figs. 1 and 2). This delayed elution of polar probes is in agreement with the literature for cellulose.17

thumbnail image

Figure 1. Detector response for injection of ethylacetate into Avicel (500 mg) sample at flow rate 15 mL/min. FID, flame ionization detector.

Download figure to PowerPoint

thumbnail image

Figure 2. Detector response for injection of ethylacetate into Methocel K4M (500 mg) sample at flow rate 15 mL/min. FID, flame ionization detector.

Download figure to PowerPoint

Elution efficacy was calculated from the integration of peak area (from the start of a peak to 60 min, because 60 min was the time when the detector response returned to an apparent baseline) and this was found to almost always be complete for nonpolar probes, but incomplete for polar probes (Tables 7–10). It can be seen that the elution efficacy of polar probes was very low for HPMC samples (Table 7 for Methocel and Table 8 for Pharmacoat). This retention of probes within the column meant that reproducibility of these results would not be good unless a very long time was left between the injections, to allow for the removal of retained probes. As might be expected, elution yield increased with flow rate increases (see, for example, the retention of ethyl acetate on Methocel 500 mg at flow rates of 15, 25, and 35 mL/min, giving probe retention of 50, 68, and 75%, respectively; Table 7). From this it can be concluded that better results are obtained at higher flow rates; however, high flow rate can cause big pressure decreases through the column and small differences between retention times of the probes and methane, so an appropriate compromise must be found. Additionally, specific interaction of polar probes with HPMC decreased with increase of flow rate (Tables 1 and 2).

Table 8. The Percentage of Probe Molecules Eluted through the Column as Compared with the Amount of Probes Injected: Pharmacoat Sample
Mass (mg)Flow (mL/min)Yield
Table 9. The Percentage of Probe Molecules Eluted through the Column as Compared with the Amount of Probes Injected: Avicel Sample
Mass (mg)Flow (mL/min)Yield
Table 10. The Percentage of Probe Molecules Eluted through the Column as Compared with the Amount of Probes Injected: Magnesium Stearate Sample
Mass (mg)Flow (mL/min)Yield

The retention of probes on microcrystalline cellulose (Table 9, Avicel PH101) was much less significant than for HPMC, and the probe retention peaks were symmetrical for the Avicel samples. This reflects the different structures of amorphous (HPMC) and the more complex microcrystalline and porous Avicel PH101. This indicates that the study of probe diffusion and retention can be potentially useful in understanding the different natures/packing morphologies of cellulosic materials.

Clearly, when probe is absorbing into samples or being retained on the column, the peak deviates from the expected Gaussian, and shows tailing. However, symmetrical retention peaks were observed for polar probes on magnesium stearate where specific interactions are much smaller in comparison to other powders (Tables 1–4). However, surprisingly, two probes (nonane and ethylacetate) did show some retention on the magnesium stearate column when the powder mass was high. Although it is reasonable for alkanes to have favorable interactions with the hydrophobic magnesium stearate, it is unclear why nonane alone should be retained.

Acid-Base Nature of the Surfaces

Despite the difficulties encountered with partial retention of polar probes, Tables 11–14 show calculated acid and base properties of analyzed powders. They were calculated from the specific interaction of three vapors (acetone, chloroform, and ethyl acetate) with the powder surfaces. With both HPMC samples and Avicel, the linear correlation of ΔGsp/AN* versus DN/AN* was good. Results for Methocel K4M (Table 11) and Pharmacoat 606 (Table 12) show that acid and base properties are affected by changes of flow rate and mass of the sample. For example, in Table 11 it can be seen that Kd/Ka can vary between 1.02 and 1.36. This is very significant, but it is unclear whether this is an artefact of the retention of vapors or due to access of more of the material as a consequence of diffusion. In any respect, it shows that workers must take great care with flow rates and sample mass in the column when considering polar probes that are able to partition into the sample. Despite the fact that the acid/base terms are affected by changes in sample mass and flow rate, it is also true that the Kd/Ka ratio for Pharmacoat is always higher than that for the other HPMC sample Methocel (Tables 11 and 12); consequently, even with the experimental impact of mass and flow rate, the IGC is able to differentiate between these samples. It can be concluded that Pharmacoat 606 is more basic than Methocel, whereas acid properties of both samples are similar. For Pharmacoat, the data have also been replotted by using the center of mass of the retention peak rather than the peak height (also in Table 12); whereas the data are slightly different when using the two calculation methods, the rank order is much the same.

Table 11. Acid-Base Properties for Methocel
Mass (mg)Flow (mg/mL)KdKaraKd/Ka
  • a

    Regression of plot used to determine Kd and Ka.

Table 12. Acid-Base Properties for Pharmacoata
Mass (mg)Flow (mL/min)KdKarbKd/Ka
  • a

    Numbers in square brackets are those determined using the center of mass of the retention peak; other numbers are from the use of peak height of the retention peak.

  • b

    Regression coefficient of line used to calculate Kd and Ka.

25071.800 [1.826]0.982 [0.967]0.997 [0.998]1.83 [1.89]
250111.549 [1.574]0.934 [0.920]0.998 [0.999]1.66 [1.71]
250151.394 [1.420]0.890 [0.876]0.999 [0.993]1.57 [1.62]
Table 13. Acid-Base Properties for Magnesium Stearate
Mass (mg)Flow (mL/min)KdKar
Table 14. Acid-Base Properties for Acyclovir
Mass (mg)Flow (mL/min)KdKar

Avicel showed lower values for both acid and basic properties than HPMC samples and magnesium stearate had the lowest values of acid/base properties among all powders analyzed. Both of these samples had negative values for Kd. Chloroform interacts with these surfaces very weakly and the standard error of measurement influences the results to a much higher degree than with other samples and it may be this error that gives rise to values being negative.

Acyclovir had dramatically different acid/base behavior when compared with the other powders, being the most acidic surface.

Despite the difficulties encountered with the retention of polar probes, there is clearly greater discrimination between the samples when considering acid-base rather than just dispersive nature.


  1. Top of page
  2. Abstract

The use of IGC in the pharmaceutical industry is growing at a very rapid rate; however, there is a great deal of research still to be performed on the fundamentals of this technique. Materials with very different properties (e.g., magnesium stearate and celluloses) have been shown to have very similar dispersive surface energies when measured at infinite dilution. It is important to develop a full understanding of exactly what is being measured and what the meaning is for a dispersive energy at infinite dilution.

The use of polar probes will become more important in order to differentiate and understand the surface natures of powders. From this study, it is clear that retention of polar probes, especially for amorphous solids, but also for fine materials (magnesium stearate in this instance), gives rise to certain problems that require a great deal of care in the use of the instrument. However, the ways in which polar probes can interact with the powders does provide hope that the study of diffusion into amorphous samples could be a very valuable tool. Furthermore, the derived acid/base data are clearly able to differentiate between the samples.

It is clear that IGC has enormous promise as a characterization tool, and that the range of uses has yet to be fully mapped out. However, when workers use and publish with this technique, they must be fully aware of the potential difficulties as well as the potential benefits. In summary, IGC can be used appropriately for predominantly crystalline materials in order to assess nonpolar surface energy; however (at least at infinite dilution), the values obtained are sometimes surprisingly more similar to one another than may be expected for very similar materials. For amorphous materials (especially) when using polar probes, there are clear problems with respect to absorption and retention of the probe. This makes conventional surface energy determination problematic; however, it may open a way to study the relative morphologies of the materials through diffusion of the probes.


  1. Top of page
  2. Abstract
  • 1
    Buckton G. 1997. Characterization of small changes in the physical properties of powders of significance for dry powder inhaler formulations. Adv Drug Deliv Rev 26: 1727.
  • 2
    Ticehurst MD, York P, Rowe RC, Dwivwdi SK. 1994. Determination of the surface properties of two batches of salbutamol sulphate by inverse gas chromatography. Int J Pharm 111: 241249.
  • 3
    York P, Ticehurst MD, Osborn JC, Roberts RJ, Rowe RC. 1988. Characterization of the surface energetics of milled DL propranolol hydrochloride using inverse gas chromatography and molecular modelling. Int J Pharm 174: 179186.
  • 4
    Feeley JC, York P, Sumby BS, Dicks H. 1988. Determination of surface properties and flow characteristics of salbutamol sulphate, before and after micronisation. Int J Pharm 172: 8996.
  • 5
    Grimsey IM, Sunkersett M, Osborn JC, York P, Rowe RC. 1999. Interpretation of the differences in the surface energetics of two optical forms of mannitol by inverse gas chromatography and molecular modelling. Int J Pharm 191: 4350.
  • 6
    Sunskersett MR, Grimsey IM, Doughty SW, Osborn JC, York P, Rowe RC. 2001. The changes in surface energetics with relative humidity of carbamazepine and paracetamol as measured by inverse gas chromatography. Eur J Pharm Sci 13: 219225.
  • 7
    Newell HE, Buckton G, Butler DA, Thielmann F, Williams DR. 2001. The use of inverse phase gas chromatography to study the change of surface energy of amorphous lactose as a function of relative humidity and the processes of collapse and crystallisation. Int J Pharm 217: 4546.
  • 8
    Fowkes FM. 1964. Attractive forces at interfaces. Ind Eng Chem 56: 4052.
  • 9
    Saint Flour C, Papirer E. 1982. Gas solid chromatography: A method of measuring surface free energy characteristics of short glass fibres. II. Through retention volumes measured near zero surface coverage. Ind Eng Chem Process Res Dev 21: 666.
  • 10
    Schultz J, Lavielle L, Martin C. 1987. The role of the interface in carbon fibre-epoxy composites. J Adhes 23: 4560.
  • 11
    Gutmann V. 1978. The donor-acceptor approach to molecular interactions. New York: Plenum, Chapter 2.
  • 12
    Riddle FL, Fowkes FM. 1990. Spectral shifts in acid-base chemistry. I. Van der Waals contributions to acceptor numbers. J Am Chem Soc 112: 32593264.
  • 13
    Dove JW, Buckton G, Doherty C. 1996. A comparison of two contact angle measurement methods and inverse gas chromatography to assess the surface energies of theophylline and caffeine. Int J Pharm 138: 199206.
  • 14
    Ahfat MN, Buckton G, Burrows R, Ticehurst MD. 2000. An exploration of inter-relationships between contact angle, inverse gas chromatography, and triboelectric charging data. Eur J Pharm Sci 9: 271276.
  • 15
    Bailey RA, Persaud KC. 1988. Application of inverse gas chromatography to characterization of a polypyrrole surface. Anal Chim Acta 363: 147156.
  • 16
    Ticehurst MD, York P, Rowe RC, Dwivwdi SK. 1996. Characterization of the surface properties of α-lactose monohydrate with inverse gas chromatography, used to detect batch variation. Int J Pharm 141: 9399.
  • 17
    Czeremuszkin G, Mukhopadhyay P, Sapieha S. 1997. Elution behavior of chemically different probes on the evaluation of surface properties of cellulose by inverse gas chromatography. J Colloid Interface Sci 194(1): 127137.