2013

http://informahealthcare.com/ddiISSN: 0363-9045 (print), 1520-5762 (electronic)

Drug Dev Ind Pharm, Early Online: 1–6! 2013 Informa Healthcare USA, Inc. DOI: 10.3109/03639045.2013.795581

ORIGINAL ARTICLE

Evaluation of risk and benefit in thermal effusivity sensor for monitoringlubrication process in pharmaceutical product manufacturing

Jumpei Uchiyama1, Yoshiteru Kato1, and Yoshifumi Uemoto2

1Manufacturing Department Kawashima Plant and 2Global Formulation Japan Pharmaceutical Science Technology, Eisai Co., Ltd., Kakamigahara,

Gifu, Japan

Abstract

In the process design of tablet manufacturing, understanding and control of the lubricationprocess is important from various viewpoints. A detailed analysis of thermal effusivity data inthe lubrication process was conducted in this study. In addition, we evaluated the risk andbenefit in the lubrication process by a detailed investigation. It was found that monitoring ofthermal effusivity detected mainly the physical change of bulk density, which was changed bydispersal of the lubricant and the coating powder particle by the lubricant. The monitoring ofthermal effusivity was almost the monitoring of bulk density, thermal effusivity could have ahigh correlation with tablet hardness. Moreover, as thermal effusivity sensor could detect notonly the change of the conventional bulk density but also the fractional change of thermalconductivity and thermal capacity, two-phase progress of lubrication process could berevealed. However, each contribution of density, thermal conductivity, or heat capacity tothermal effusivity has the risk of fluctuation by formulation. After carefully considering thechange factor with the risk to be changed by formulation, thermal effusivity sensor can be auseful tool for monitoring as process analytical technology, estimating tablet hardness andinvestigating the detailed mechanism of the lubrication process.

Keywords

Magnesium stearate, mechanism oflubrication process, monitoring lubricationprocess, process analytical technology,tablet hardness, thermal effusivity

History

Received 11 February 2013Revised 7 April 2013Accepted 9 April 2013Published online 21 May 2013

Introduction

A compressed tablet is the most popularly used dosage form.However, manufacturing of tablets is a complex process, becauseonly a few materials have physical properties which are requisitefor the manufacturing of tablets of adequate quality. Therefore,some treatment and incorporation of excipients in the formulationis commonly needed. The well-thought-out development offormulation and manufacturing process is needed for manufactur-ing of tablets of sufficient quality.

The lubricant is one of the most important excipients informulation. The lubricant has the ability to improve the lubricityand flowability of the powder. An adequate concentration and adispersible state of the lubricant are able to get the appropriatecompressed tablet of sufficient quality, and make high-speedtableting possible. The dispersible state is a result of the blendingtime of the lubricant. However, the lubricant often causes sometrouble during tablet manufacturing and has some negative impacton tablet properties1–3.

Although lubricants such as magnesium stearate is often usedin very low concentrations (0.2–2.0%), the optimization andverification of concentration and blending condition are required

for tablets with sufficient quality. If the concentration is too low orif the blending time is inadequate, the powder sticks on the tabletpress punches, or the tablets bind within the die cavity, resultingin excessive ejection forces and broken tablets. The tablets mightalso experience picking or sticking as a result of adhesion topunch faces. Blended products with a high concentration oflubricant or long blended time can often cause other problems,resulting in complete particle coating. These problems includeless blend compressibility, deficient tablet hardness, extension oftablet disintegration times, and delay in dissolution rates.Therefore, the control of the lubrication process is critical inaspects of various effects on tablets with sufficient quality.

Lubrication process is a very simple blending operation, buttablet physical properties are greatly affected by formulation,blending time and so on. In addition, a detailed mechanism of thelubrication process is still unknown. For the decision of endpointin the lubrication process, it is generally necessary to evaluatepowders and tablet properties as a function of blending timecourse, and optimal blending condition is verified. In the case of achange in the lubrication conditions such as the changes ofproduction scale or blender-type, it is a time-consuming processto perform the optimization study and to re-verify the blendingcondition.

The guidelines for process analytical technology (PAT) issued4

by the Food and Drug Administration (FDA) promote online,real-time analyses as a tool to monitor and control product quality.Online analyses can reduce or eliminate reworking batches,

Address for correspondence: Jumpei Uchiyama, Manufacturing Depart-ment Kawashima Plant, Eisai Co., Ltd., 1, Kawashimatakehaya,Kakamigahara, Gifu 501-6195, Japan. E-mail: j-uchiyama@hhc.eisai.co.jp

Dru

g D

evel

opm

ent a

nd I

ndus

tria

l Pha

rmac

y D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y V

ande

rbilt

Uni

vers

ity o

n 05

/24/

13Fo

r pe

rson

al u

se o

nly.

decrease the frequency of finished product testing, increasemanufacturing efficiency, and ensure product quality throughoutthe manufacturing process.

Recently various evaluations in the lubrication process aretested to perform analyses by PAT tool like near infrared, NIR3,5,6,thermal effusivity7–13 and so on. For example, by using the NIRspectra of raw blended powder, its calibration models were able toevaluate not only tablet hardness but also the CV of tablet weight.It was suggested that the tablet hardness and CV of tablet weightcould quickly and nondestructively be predicted before compres-sion. However, in the NIR method, analysis and calibration ofchemometric data require tremendous amounts of time for thedevelopment of a standard curve in each case, and NIR spectradata are intricately affected by the physical property such asparticle size. In addition, the lubrication process is not only theblending of very little constituent but also physical change inparticles coated by lubricant. Although chemometric NIR methodis suitable for a quantitative analysis of a certain constituent at themicro level, it is unsuitable for detecting very little constituent andphysical change at the macro level. Terashita reported12 that endpoints in lubrication using NIR method were different from theend point of thermal effusivity. Therefore, it is assumed thatchemometric NIR method is unsuitable for the monitoring oflubrication process, and it is not widely prevalent technology nowin lubrication process7.

However, use of the thermal effusivity sensor is a nondestruc-tive method, and pretreatment and chemometric data analysis isnot required like the NIR method. From this advantage, thethermal effusivity sensor was studied not only in the lubricationprocess but also in other processes of pharmaceutical productsmanufacturing. It was applied to blend the uniformity process,roller compaction process, drying process, and so on. Thermaleffusivity values indicated that the blend of the eight componentswas sensitive to uniformity8. Thermal effusivity had a strongcorrelation with the physical properties of compacted ribbons,which could be used to monitor these properties9. The sensor wasused in the powder drying process for the purpose of its sensitivity

to the moisture content10.Thermal effusivity sensor is attracting attention in lubrication

process monitoring. In previous reports, thermal effusivityincreased over time in the lubrication process7, the sensor wasable to monitor and detect the difference in the additive amount ofmagnesium stearate in the lubrication process11. Moreover, it hashigh correlation with tablet hardness in the lubrication process12.However, although tablet hardness was able to be estimated bymeasuring the thermal effusivity, it is unclear as to what thermaleffusivity sensor detects and whether the thermal effusivity sensor

is an absolutely useful tool as a process controller in any caseduring the lubrication process.

In this study, we closely examine previous reports7,11;moreover, we evaluate the risk and benefit in the use of thermaleffusivity sensor for monitoring lubrication process. We tried tounderstand what is going on in the lubrication process fromthermal effusivity sensor. It was investigated further as to whythermal effusivity has a high correlation with tablet hardness andwhat thermal effusivity sensor detects in the lubrication process.

Materials and methods

Materials

Lactose (Tablettose 80, Meggle G.m.b.H., Wasserburg,Germany), microcrystalline cellulose (Avicel, PH-101, AsahiChemical Industry Co. Ltd., Tokyo, Japan) and corn starch (NihonShokuhin Kako Co., Ltd., Tokyo, Japan) were used as fillers,materials of blend before lubrication. Magnesium stearate(Mallinckrodt Japan Co. Ltd., Fukuoka, Japan) was used as alubricant.

Lubrication blending operation

First, lactose 80 (11.96 kg), microcrystalline cellulose (6.00 kg)and cornstarch (2.00 kg) were blended before the lubricationoperation using a tumbler mixer (TM-50, Showa KagakukikaiKosakusho Co., Ltd., Japan) at a speed of 16.4 rpm for 30 min.Second, magnesium stearate (0.04 kg) was added into the powdermixtures. These were blended under the same rotation speed. It was70.18% of the total volume of blender. During the lubricationprocess, the blender was stopped at predefined intervals, andblending of the powder in each time was taken for measuringthermal effusivity (10 g) and tableting (50 g). This sample wastaken from under 3 cm in the center of the powder layer in a blenderas viewed from above by using a 50 ml cup. The interval time andmeasured point is shown as below. The point where the thermaleffusivity did not almost change was considered the endpoint.

Tableting operation

The sampled powder in each lubrication time was tableted by aneccentric tableting machine (N-30EX, Okada seiko Co., Ltd.,Tokyo, Japan) at a compression force of 10 kN. The diameter ofthe punch was 8.0 mm, and the weight of the tablet was 200 mg.

Measurement of thermal effusivity

The thermal effusivity sensor (Thermal Effusivity Sensor MathisESP-04, C-therm Technologies, Canada) was used. The powder in

2 J. Uchiyama et al. Drug Dev Ind Pharm, Early Online: 1–6

Dru

g D

evel

opm

ent a

nd I

ndus

tria

l Pha

rmac

y D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y V

ande

rbilt

Uni

vers

ity o

n 05

/24/

13Fo

r pe

rson

al u

se o

nly.

each lubrication time (10 g) was delicately put into an offline-kit.The cooling time was 60 seconds and the measuring time was 1.5seconds. The average of three time measurements was calculated.

Determination of bulk density

For samples taken out from the blender, the bulk density wasdetermined by pouring delicately presieved (10 mesh) powderblend into a graduated cylinder via a large funnel and bymeasuring the volume and weight14. The volume, diameter andheight of the cylinder were 30 ml, 21.5 mm and 83.0 mm,respectively. The distance between the cylinder and the funnelwas 15.0 mm. The bulk density is calculated from Equation (1).This filling method to measure bulk density was in pretty muchthe same as thermal effusivity by off-line kit.

Bulk density ¼ bulk volume

weight the powder blendð1Þ

Determination of tablet hardness

Tablet hardness was measured by the tablet hardness tester(Fujiwara Factory Ltd., Japan). The average hardness of 10 tabletswas calculated.

Results and discussion

Determination of thermal effusivity during the lubricationprocess

The change of thermal effusivity with the number of accumulatedrotations of the blender is shown in Figure 1. Before starting thelubrication process, thermal effusivity was 240 Ws1/2/m2K. Afterstarting the lubrication process, thermal effusivity rose linearlywith the number of accumulated rotations to 2000 rotations andreached a constant, 325 Ws1/2/m2K. Each point was triplicated andthe variation was small. Roy11 reported that the thermal effusivityrose with blending time from 270 Ws1/2/m2K to 340 Ws1/2/m2Kduring the lubrication process, in a mixture of lactose, micro-crystalline cellulose and magnesium stearate (2.0%, w/w) blendedin a 325 l bin container. In addition, it was the similar result whenmagnesium stearate (1.0, 2.0, and 4.0%, w/w) was blended in a1-cubic-foot V-shell blender and 2-cubic-foot V-shell blender.Terashita12 also reported that thermal effusivity rose withblending time from 270 Ws1/2/m2K to 290 Ws1/2/m2K during thelubrication process, in a mixture of acetaminophen, lactose andmagnesium stearate (1.0%, w/w) blended in a hi-speed mixer.Although the type of blender and formulation were different fromthis study, the rising trend of thermal effusivity in the lubricationprocess showed similar profiles. Moreover, even though they used

more than 1.0% concentration of magnesium stearate as alubricant, it was found that the thermal effusivity sensor coulddetect change of thermal effusivity in case of even very lowconcentrations (0.2%, w/w) of lubricant in this study. Therefore,the thermal effusivity sensor could monitor lubrication processesat various concentration of magnesium stearate and at variousblender types.

Relationship between thermal effusivity and tablethardness

On the impact on tablet properties, understanding of the conditionin the lubrication progress is requisite for the manufacturing ofsufficient tablet quality. In the lubrication process, dispersal andcoating of lubricant occurred intricately. Also the lubricationprocess is not only the blending of very little constituent in thelubricant but also a physical change in particles coated by thelubricant. It was necessary to investigate the relationship betweenthe condition in the lubrication progress and tablet properties.

Tablet hardness is one of the most important tablets properties.In the lubrication process, tablet hardness was susceptiblychanged by the change of lubrication condition, formulation andso on. Wang and others reported that tablet hardness correlatedwell with tablet disintegration times, dissolution rates and so on,in the lubrication process1–3. Therefore, we chose tablet hardnessas a representative of tablet properties.

The relationship between thermal effusivity and tablet hard-ness is shown in Figure 2. As thermal effusivity increases from240 to 325 Ws1/2/m2K, tablet hardness decreased linearly from150 to 75 N. It was confirmed that there was a strong negativecorrelation between thermal effusivity and tablet hardness (thevalue of correlation coefficient was 0.9893). Although formula-tion and blender type were different on this study, Terashita alsoreported12 that tablet hardness declined from 105 to 85 N withthermal effusivity rising from 270 to 290 Ws1/2/m2K (four pointsduring lubrication process). Therefore, once the relationshipbetween thermal effusivity and tablet hardness is understood, it iseasy to get target tablet hardness in each lubrication process. Forexample, if tablet hardness of 100 N is needed, the lubricationprocess should stop at 300 Ws1/2/m2K in this study. It was shownthat thermal effusivity sensor might be very useful tool formonitoring the lubrication process and estimating tablet hardness.

In common with previous reports12, it was confirmed thatusing thermal effusivity could estimate tablet hardness in thisstudy. However, thermal relationship between effusivity and tablethardness was not researched in detail. The report of Terashita12

that shows similar findings to this study did not also discuss thisassignment. Also, Yonemochi7 reported that thermal effusivity

Figure 2. Relationship of thermal effusivity to tablet hardness throughoutlubrication process.Figure 1. Thermal effusivity data throughout lubrication process.

DOI: 10.3109/03639045.2013.795581 Thermal effusivity sensor for monitoring lubrication process 3

Dru

g D

evel

opm

ent a

nd I

ndus

tria

l Pha

rmac

y D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y V

ande

rbilt

Uni

vers

ity o

n 05

/24/

13Fo

r pe

rson

al u

se o

nly.

has some relationship with bulk density, but it was yet to beidentified by sufficient data.

The reason why thermal effusivity was high correlation (thevalue of correlation coefficient was 0.9893) to tablet hardnessremains to be elucidated. Sufficiently detailed relationshipbetween thermal effusivity and tablet hardness was not reportedpreviously. This relationship had to be circumstantiallyinvestigated.

Relationship between thermal effusivity and bulk density

Shah reported that the change of powder bulk density was quitecomparable with the change of tablet hardness during lubricationprocess1, suggesting that the influence of blending progress onboth properties was a manifestation of the enlarged surface withseparation or the greater surface coverage by the lubricant onblending. Generally, it is well-known that the relationshipbetween tablet hardness and bulk density has high correlation.

The change of bulk density with the number of accumulatedrotations is shown in Figure 3. Before starting the lubricationprocess, the bulk density was 490 kg/m3. After starting thelubrication process, a lag was observed and the bulk densityincreased linearly with the number of accumulated rotations to2500 rotations and reached a constant 650 kg/m3. The rising curveof bulk density was similar to the rising curve of thermaleffusivity, but the constant point reached in bulk density (2500rotations) as shown in Figure 3 was later than in thermal effusivity(2000 rotations) as shown in Figure 1. From both similar profiles,thermal effusivity appeared to be associated with bulk density.

Thermal effusivity is an indicator of transferring efficiencywhich thermal energy passes through contacting surface when twoobjects at different temperatures touch. The thermal effusivity iscalculated from Equation (2).

E ¼ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffik � � � Cp

pð2Þ

E, k, �, and Cp are thermal effusivity (Ws1/2/m2K), thermalconductivity (W/m K), bulk density (kg/m3), and heat capacity (J/kg K), respectively. Thermal effusivity sensor heats sample fromits surface, measures its surface temperature, and calculatesthermal effusivity from the grade change of temperature. Thegrade change of temperature is inversely related to absorptionthermal capacity of sample, so the higher grade is, the lowerthermal effusivity is. Indeed, the sensor measured thermaleffusivity of powder and interparticle air in parallel. However,the thermal effusivity of solids and gases substantially differsfrom the thermal effusivity of powders.

The relationship between the square root of bulk density, SBD,and thermal effusivity is shown in Figure 4. The figure shows atwo-phase process. When SBD was below 22.5 kg1/2/m3/2, therewas no clear relationship between SBD and thermal effusivity. Onthe other hand, when SBD was over 22.5 kg1/2/m3/2, thermaleffusivity had high correlation with SBD (the value of correlationcoefficient was 0.9760).

The change of powder bulk density was almost the same asthe change of tablet hardness. It was suggested that the changeof both properties was a phenomenon of the enlarged surface ofseparation or greater surface coverage by the lubricant onblending1. Because the monitoring of thermal effusivity wasalmost equal to the monitoring of bulk density as shown inFigure 4, the thermal effusivity sensor could have a highcorrelation with tablet hardness (the value of correlation coeffi-cient was 0.9893) as shown in Figure 2. The bulk density cannotbe measured at in-line during the lubrication process directly.However, if the thermal effusivity sensor is installed in theblender as per the report of Roy11, the bulk density and tablet

hardness can be estimated by monitoring the thermal effusivityin-line. It was thought that the thermal effusivity sensor is a usefultool as PAT for lubrication process, because the change of thermaleffusivity value could nearly identify with the change of bulkdensity.

The thermal effusivity sensor might be adequate to estimatetablet hardness on the practical side as PAT tool in this study.

Contribution of density, thermal conductivity or heatcapacity to thermal effusivity

Although the thermal effusivity sensor could have high correl-ation with tablet hardness (the value of correlation coefficient was0.9893), it remains to be elucidated as to whether or not the use ofthermal effusivity sensor as a process control tool is absolutelyuseful in any case during lubrication process. Also the effect ofthermal conductivity and heat capacity other than bulk density onthermal effusivity has not been discussed. Therefore, for having amuch better understanding and for evaluating the risk and benefitof thermal effusivity sensor in the lubrication process, thecontribution of density, thermal conductivity or heat capacity tothermal effusivity was investigated in detail.

Using the result of the change of thermal effusivity and bulkdensity, it was investigated whether thermal effusivity sensordetects only bulk density. Also, the contribution of density,thermal conductivity or heat capacity to thermal effusivity wasinvestigated.

Relationships between the number of accumulated rotationsand SBD, the square root of thermal conductivity multiplied by

Figure 4. Relationship of thermal effusivity to the square root of bulkdensity (SBD) throughout lubrication process.

Figure 3. Bulk density data throughout lubrication process.

4 J. Uchiyama et al. Drug Dev Ind Pharm, Early Online: 1–6

Dru

g D

evel

opm

ent a

nd I

ndus

tria

l Pha

rmac

y D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y V

ande

rbilt

Uni

vers

ity o

n 05

/24/

13Fo

r pe

rson

al u

se o

nly.

thermal capacity, STT, are shown in Figure 5. STT was calculatedfrom both SBD and thermal effusivity as shown in Equation (3).

ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiðk � CpÞ

q¼ Effiffiffiffiffiffiffi

�ð Þp ð3Þ

Before starting the lubrication process, SBD was 22.0 kg1/2/m3/

2, and STT was 11.0 Ws1/2/m1/2kg1/2K. After starting lubricationprocess, SBD increased with the number of accumulated rotationsto 25.5 kg1/2/m3/2 at 3000 rotations. Meanwhile, STT increasedwith the number of accumulated rotations to 800 rotations andreached a constant 13.0 Ws1/2/m1/2kg1/2K. The change of thermaleffusivity after 800 rotations showed the change of SBD directly,because STT after 800 rotations was a constant. This is the reasonwhy thermal effusivity had a high correlation with SBD (the valueof correlation coefficient was 0.9760) when SBD was over22.5 kg1/2/m3/2, as shown in Figure 4. Therefore, when SBD wasbelow 22.5 kg1/2/m3/2, there was no clear relationship betweenSBD and effusivity, before 800 rotations (first phase). On theother hand, when SBD was over 22.5 kg1/2/m3/2, there was stablehigh relationship between SBD and thermal effusivity, after 800rotations (second phase).

The value of SBD was about twice as great as the constantvalue of STT. The contribution of SBD was bigger than thecontribution of STT before 800 rotations. Therefore, it wasthought that the main factor affecting the change of thermaleffusivity was bulk density more than thermal conductivity andthermal capacity. However, thermal conductivity and thermalcapacity did not even have a small effect on the change of thermal

effusivity in the case of this study. The thermal effusivity sensorcould detect not only the bulk density but also fractional thermalconductivity and thermal capacity. From this result, detailedlubrication progress could be predicted as follows.

In the lubrication process, dispersal and coating of magnesiumstearate as a lubricant is intricately going on. Roblot15 reportedthat the distribution of magnesium stearate on the surface ofgranules is investigated using an electron microscope andmicroanalysis, and the preferential location of the magnesiumstearate in cavities, and the regularization of the surface providedby the lubricant. Therefore, each main factor in the two-phaseprocess was estimated as shown below. About the main factor ofthe first phase, it was assumed that a powdery overall thermo-chemical value was changed by dispersion of magnesium stearate.In addition, it was completed at 800 rotations, and was supposedthat magnesium stearate reached a nearly homogeneous disper-sion state. This was because the change of STT was sharper thanthe change of SBD until 800 rotations. About the main factor ofsecond phase, it was thought that the bulk density was changed bywhat magnesium stearate coated in the powder particle. This wasbecause the change of thermal effusivity showed the change ofbulk density directly after 800 rotations.

Terashita12 investigated lubrication process by using NIR andthermal effusivity. He reported that both end points of lubricationprocess were different results. The end point of NIR method(1 min) was earlier than the end point of thermal effusivitymethod (between 2 and 3 min). From this result about the differentresults of both end points of lubrication process by analyzing NIRand thermal effusivity, it could be assumed as follows. Monitoringof NIR detected mainly dispersal of the lubricant, because thisapproach was only analysed by uniformity of a component. On theother hand, monitoring of thermal effusivity detected mainly aphysical change of bulk density which was changed by dispersalof lubricant and coating powder particle by lubricant. However,investigating this speculative theory is an issue in the future.

In this study, although thermal effusivity consists of thermalconductivity, bulk density and heat capacity, STT made a smallcontribution to thermal effusivity compared with SBD. Though,by taking into account not only the big contribution of bulkdensity but also the small contribution of thermal conductivityand heat capacity, the elucidation of detailed mechanism inlubrication process was able to be expected.

However, the percentage of contribution for each density,thermal conductivity, or heat capacity to thermal effusivity hasthe potentiality to be fluctuated by formulation. Before usingthermal effusivity as a PAT tool for determination of lubricationend point, each percentage in the contribution with the potenti-ality to be changed by formulation should be considered. If STTbecomes a big contribution to thermal effusivity, more thanin this study, there is the risk that the relationship between tablethardness and thermal effusivity has no high correlation.Therefore, by considering the change factor of thermal effusivitysensor in the lubrication process more carefully, it was able toreveal that the thermal effusivity sensor could be a useful tool formonitoring as PAT, estimation of tablet hardness, and investiga-tion of detailed mechanism, in lubrication process.

Because of detailed analysis to thermal effusivity data by SBDand STT, the thermal effuisvity sensor is able to monitor thelubrication process as PAT, and has the potential to reveal themechanism of lubrication process such as two-phase process in anew and different way unlike using NIR.

Conclusion

Although the thermal effusivity sensor could have a highcorrelation with tablet hardness, it remains to be elucidated that

Figure 5. The square root of bulk density (SBD) and the square root ofthe thermal conductivity multiplied by thermal capacity (STT) throughoutthe lubrication process SBD and STT.

DOI: 10.3109/03639045.2013.795581 Thermal effusivity sensor for monitoring lubrication process 5

Dru

g D

evel

opm

ent a

nd I

ndus

tria

l Pha

rmac

y D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y V

ande

rbilt

Uni

vers

ity o

n 05

/24/

13Fo

r pe

rson

al u

se o

nly.

using the thermal effusivity sensor as process control tool isabsolutely useful in any case during lubrication process. Also theeffect of thermal conductivity and heat capacity other than bulkdensity to thermal effusivity is not discussed.

Therefore, for a much better understanding and for evaluatingthe risk and benefit of the thermal effusivity sensor in thelubrication process, the contribution of density, thermal conduct-ivity or heat capacity to thermal effusivity was investigated indetail.

In the result, thermal effusivity sensor could be a useful toolfor estimating tablet hardness in the lubrication process. Becausethe monitoring of thermal effusivity was almost the monitoring ofbulk density, the thermal effusivity sensor was able to have a highcorrelation with tablet hardness. Moreover, the two-phase pro-gress of the lubrication process was revealed in this study becausethe thermal effusivity sensor could detect not only a fractionalchange of the conventional bulk density but also a fractionalchange of thermal conductivity and thermal capacity in case ofthis study. The monitoring of more detailed lubrication processwas possible by using thermal effusivity sensor.

However, each contribution of density, thermal conductivity orheat capacity to thermal effusivity has the potentiality to fluctuateby formulation. Using the thermal effusivity sensor, the changefactor of thermal effusivity sensor in the lubrication process has tobe considered more carefully. After carefully considering thechange factor, the thermal effusivity sensor could be a useful toolfor monitoring as PAT, estimating of tablet hardness, andinvestigating a detailed mechanism, in lubrication process.Moreover, detailed analysis of thermal effusivity data by bulkdensity, thermal conductivity and thermal capacity revealed aclose mechanism of lubrication process such as two-phaseprocess.

Declaration of interest

The authors report no conflicts of interest. The authors alone areresponsible for the content and writing of this article.

References

1. Shah AC, Mlodozeniec AR. Mechanism of surface lubrication:influence of duration of lubricant-excipient mixing on processing

characteristics of powders and properties of compressed tablets. JPharm Sci 1977;66:1377–82.

2. Wang J, Wen H, Desai D. Lubrication in tablet formulations. Eur JPharm Biopharm 2010;75:1–15.

3. Abe H, Otsuka M. Effects of lubricant-mixing time on prolongationof dissolution time and its prediction by measuring near infraredspectra from tablets. Drug Dev Ind Pharm 2012;38:412–19.

4. FDA (2004). Guidance for Industry, PAT – A Framework forinnovative pharmaceutical development, manufacturing, and qualityassurance. Available from: http://www.fda.gov/downloads/Drugs/GuidanceComplianceRegulatoryInformation/Guidances/UCM070305.pdf [last accessed Sep 2012].

5. Otsuka M, Yamane I. Prediction of tablet properties based on nearinfrared spectra of raw mixed powders by chemometrics: scale-upfactor of blending and tableting processes. J Pharm Sci 2009;98:4296–308.

6. Terashita K, Sato T. Feasibility study: real-time monitoring ofblending end point of magnesium stearate by AOTF-NIR spectrom-eter. Pharm Tech Japan 2008;24:613–20.

7. Yonemochi E. Application of thermal effusivity sensor to evaluatethe manufacturing and quality of pharmaceuticals. Netsu Sokutei2008;35:180–4.

8. Mathews L, Chandler C, Dipali S, et al. Monitoring blend uniformitywith effusivity. Pharm Tech 2002;April:80–4.

9. Ghorab MK, Chatlapalli R, Hasan S, Nagi A. Application of thermaleffusivity as a process analytical technology tool for monitoring andcontrol of the roller compaction process. AAPS PharmSciTech2007;8:155–61.

10. Roy Y, Closs S, Mathis N, Nieves E. Thermal effusivity as a processanalytical technology: to optimize, monitor, and control fluid-beddrying. Pharm Tech 2004;November:21–8.

11. Roy Y, Mathis N, Closs S, et al. Online thermal effusivitymonitoring: a promising technique for determining when toconclude blending of magnesium stearate. Tablets and Capsules2005;3:38–47.

12. Terashita K. Meeting PAT requirements by evaluating the mixingand distribution of magnesium stearate lubricant and other compo-nents in a tablet blend using on-line analytical method part 1. PharmTech Japan 2012;28:921–6.

13. Harrisa A, Sorensen DN. Thermal conductivity testing of minimalvolumes of energetic powders. J Pyrotechnics 2007;25:49–54.

14. Kapil R, Kapoor DN, Dhawan S. Flow, compressive, andbioadhesive properties of various blends of poly(ethylene oxide).Drug Dev Ind Pharm 2010;36:45–55.

15. Roblot-Treupel L, Puisieux F. Distribution of magnesium stearate onthe surface of lubricated particles. Int J Pharm 1986;31:131–6.

6 J. Uchiyama et al. Drug Dev Ind Pharm, Early Online: 1–6

Dru

g D

evel

opm

ent a

nd I

ndus

tria

l Pha

rmac

y D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y V

ande

rbilt

Uni

vers

ity o

n 05

/24/

13Fo

r pe

rson

al u

se o

nly.