E

Reprinted From September 2004 www.tabletscapsules.com

tablet press operationPreventing and fixing weight and

hardness defects: Strategies for

production personnel

Michael D. Tousey

DI Pharma Tech

Tablet specifications are tight, and the list of possible defects islong: Variable weight, sticking, picking, black spots, streaks,capping, lamination, variable hardness, among others. Thisarticle focuses on variations in tablet weight and tablet hard-ness. It pinpoints the possible causes of these defects and offersadvice on preventing and fixing the source of the problems. Italso discusses the problems of formulations with too manyfines.

very product behaves differently on a tablet press,even if it’s the same product run on a different day. Thevariation often stems from changes in the properties ofthe raw materials—active ingredients and excipients—from batch to batch. Naturally, the goal is to minimizethese changes. Tablet press operators, however, don’thave any control over formulation and granulation. Theyhave to work with what they’re given, and their employ-ers expect them to make good tablets day in, day out.

f-Touseyart_34-43 12/1/04 11:17 AM Page 34

Tablet press adjustmentsAs the tablet press operator, your job is to make good

tablets, no matter how poorly the product works on thetablet press. Here are some things you can adjust on thetablet press to overcome weight variations. Refer toFigure 1.

Figure 1

An illustration of the compression cycle

1. Product in feeder2. Scraper3. Fill cam (partial view)4. Weight adjustment cam5. Ejection cam6. Fill position7. Weight adjustment position8. Pull-down position9. Pre-compression10. Main compression

Direction of rotation Cou

rtesy

of t

he A

mer

ican

Pha

rmac

eutic

al A

ssoc

iatio

n

1 2

3

45

6 7 8

9 10

Head pressure. This factor is related to the amount ofproduct in the overhead feed system or hopper. Themore product present, the greater the head pressure, andvice versa. When the pressure varies, so does the weightof the tablets.

Hopper level. To keep the head pressure consistent,you need to maintain the product in the hopper at a con-sistent level. Maintaining the level within a narrow rangewill reduce the potential for weight changes. Every oper-ator knows that tablets produced at the end of a runweigh less than those made at the start. That’s becausethe hopper level (and thus head pressure) decreases dur-ing the run. If you determine that variations in the hop-per level are linked to variations in tablet weight, youmust monitor the weights more often and adjust yourequipment so the level is more consistent.

Feeder clearance. Generally speaking, you should setthe clearance between the feeder and the die tableaccording to the product’s particle size. The smaller theparticle size, the smaller the clearance between the feederbase and the die table, and vice versa. That’s the generalrule, but there’s a big “if” involved here: If the powder iswaxy or very moist, a feeder that is incorrectly set maycompact, or “pack” the powder on the die table. In such

Tablet weight: Sources of variationIf you’re a tablet press operator, the exchange below

may have a familiar ring to it.Tablet press operator: It says right here that if the operator

cannot maintain target weights within acceptable qualitylimits that I must report to you, my supervisor. So whatshould I do?

Supervisor: You’re the operator. How’d you do it lasttime? It worked fine on the last batch, right?

Operator: I didn’t do anything differently. It must be theproduct. So what do you want me do?

Supervisor: Let’s pull out the checklist and review theoptions.

So the two of them scan the checklist on weight con-trol, which tells them that tablet weights are mainlyaffected by

1. Product variation2. Machine condition3. Tooling condition4. Flow of powder on the machineLet’s look at each of these factors.

Product variation. Sources of product variationinclude inconsistent powder density and particle size dis-tribution. While you may think that density variationsoriginate with your colleagues in formulation and granu-lation, density can also change on the press, oftenbecause of overfilling of the die and re-circulation of thepowder on the tablet press. The particle size distributioncan change when the product becomes “unblended” dur-ing transfer or because of static electricity. It might alsochange because the product can’t withstand the handlingand the mechanical stress it undergoes before reachingthe tablet press.

Machine condition. The problems caused by a tabletpress that is poorly prepared or operated are legion. Lookat the die table. Have you inspected its runout? The upand down motion under load on a new die table should bewithin 0.003 inch of the setting. You must also ensure thatthe pressure rolls and cams are in very good condition.

Tooling condition. There is quite a bit of informationavailable on tooling, but in a discussion of weight varia-tion, the punch working length is the property that youmust know. Working length is the key factor in howpunches affect tablet weight. New tools are made to atolerance of one-thousandth of an inch, which is equal tothe thickness of one page in this magazine. Make sure thelength of each punch is correct and identical.

Powder flow and feed-rates. Many of the defectsrelated to powder flow and feed-rates stem from the for-mulation you’re given, not the operation of the tabletpress. Since the focus of this article is resolving problemsat the tablet press, I won’t go into the many problemsthat may be related to the granulation process or the for-mulation itself except to say that too many fines is bad. Iaddress fines in more detail at the end of the article andin the sidebar on page 42.

f-Touseyart_34-43 12/1/04 11:17 AM Page 35

cases, the only wayto determine thebest setting is toexperiment with dif-ferent clearances.

Feeder speed.The feeder deliverspowder from thehopper to the dietable, and the rotat-ing paddles in thefeeder are intendedto match the pow-der flow rate to therotational speed ofthe press. Thefeeder can also un-mix blended pow-ders, and the pad-dles can breakfriable granules andcompact highlymoist powders. Thebest strategy is torun the feeder asslowly as possible while maintaining proper tabletweight. Running the feeder too fast can also cause theproduct to segregate. See Figure 2.

Fill cam. The fill cam allows the die cavity to be over-filled. The excess is then pushed up and out of the diewhen it reaches the weight adjustment cam so that thescraper blade can remove the excess as the die passes by.Most modern tablet presses include a “pull-down” posi-tion after the weight adjustment step. I discuss this fea-ture later in the arti-cle. Thin tabletsrequire a shallow fillcam, and thicktablets need a deepfill cam. It’s typicallynecessary to overfillthe cam by 10 to 30percent to maintainconsistent tabletweights.

Scraper bladecondition and ad-justment. Thescraper blade is awear part. It’s wornby die table rotationand product abra-sion. The moreabrasive the prod-uct, the faster thescraper’s edge willwear. So inspect theblade frequently and

replace it as needed.When the blade isin bad condition orset up incorrectly, itwon’t scrape the dietable clean, andyou’ll have problemswith weight control.

Re-circulation.After the scraperblade levels thepowder in the die, itguides the excesspowder into a re-circulation channel.This allows thepowder to travel onthe turret until itreturns to the feeder.If the fill cam is settoo deep, there maybe “an excess ofexcess” powder inthe re-circulationchannel. When that

happens, the powder will not flow well. It may even backup, spill, compact, or leak, all of which can reduce yieldand alter the characteristics of the re-circulated powder.Poor re-circulation produces a powder that is over-worked, compacted, unmixed, or otherwise damaged.

To see what’s happening on the press, use a strobelight. It lets you see how well the product flows, how wellthe scraper blade works, and how well the die andpunches are functioning. See Figure 3.

More on weight variation

Those are someof the machine-related variablesthat can affect tabletweight. But what ifthe weight problemis related to theproduct, not themachine? In somecases, you’ll betempted to returnthe problematicproduct to the gran-ulating department.More likely, you’llbe asked to consultyour supervisor.

Operator: So areyou going to answermy question? Whatdo I do?

Figure 2

Removing the feeder from this tablet press reveals segregated powder, which ultimately leads to weight fluctuations. Running the feeder too fast is

one cause of segregation.

Figure 3

A strobe light can reveal how well a product flows, how well the scraper bladeworks, and how well the die and punches are functioning.

f-Touseyart_34-43 12/1/04 11:17 AM Page 36

to exhaust the other options. Changing the powder den-sity is likely impossible, unless it’s a machine-induceddensity problem, as described earlier.

You can increase dwell time to increase hardness, butlonger dwell times typically require you to slow themachine, and that’s not usually an option. Instead, I rec-ommend that you first decrease tablet thickness by adjust-ing main compression. Or, if you are already at the targetweight, then decrease the thickness and increase the pres-sure at pre-compression. If you make that change andyou’re still achieving your target weight—that is, thetablet hardness increases and decreases rapidly but theweights are consistent—then you have what is known as adwell-sensitive granulation. Finally, as mentioned earlier,you must know that the tooling is in good condition. Ifthe diameter of the tooling head flat is inconsistent withinthe set of punches, then hardness will be inconsistent.

To summarize: If you have trouble controlling hard-ness, first check for consistent weights. If the weights areconsistent, then check the thickness. If the thickness isconsistent and on target, then reduce the thickness atpre-compression. Only increase the weight or reducemachine speed when you have determined that there areno other options.

A word on particle sizeI’m often asked to define the optimum particle size

distribution for making the perfect tablet. That’s a toughquestion. A particle size distribution within the rightrange produces a good-looking tablet, but the nature ofthat distribution depends on the tablet size. If the pow-der contained only large particles, like granular sugar,you’d be making a porous sugar cube. Conversely, if allthe particles were superfine, you’d have other problemswith compression. We need a range of particle sizes thatis in harmony with the tablet size.

Generally speaking, the smaller the tablet, the smallerthe particle size. Likewise, the bigger the tablet, the big-ger the particle size. Particles smaller than 200 mesh (74microns) are generally a problem for compression.Particles larger than 18 mesh (1,000 microns) can causeweight fluctuations and create a textured appearance.This is true for the most common sizes of pharmaceuticaland health-supplement tablets. Tablet shape is also a con-sideration. If the tablet has small sharp corners, then theparticles must be small enough to flow and fit into thosetight corners. However, if you’re making a round orcaplet-shaped tablet, larger particles work better.

Disintegration and dissolution times are generally betterfor tablets made with a fine dust-like powder than they arefor tablets made with large-particle powders. Finding ahappy medium—a formulation that disintegrates and dis-solves well and that also compresses well—requires sometrue formulating skills. I think that formulators too often erron the fine-particle side of the issue and don’t consider allthe issues of tablet making. In short, fines affect yields,punch lubrication, compressibility, and can cause the tabletsto cap, laminate, or show other defects. Fines also increase

Supervisor: Hmm. Isn’t the next shift going to be heresoon? Just joking!

In the real world, we have to solve the problems, notpass them along to the next shift. The first point toremember: Change one (and only one!) variable at atime. The next step is to define the extent of the prob-lem by checking the tablets closely. Is there a pattern tothe weight variation among individual tablets? Is thevariation consistent? Does it come and go? Is the trendtoward overweight or underweight tablets or does it goboth ways?

At some point you might think, “But I’ve got auto-matic weight control on my tablet press! Why is thishappening in the first place?” Well, you have to under-stand that “automatic weight control” works by allowingthe tablet press to make weight adjustments based onvariations in compression force. When the compressionforce is below the limit, an electronic signal tells thetablet press to fine-tune the weight adjustment cam toincrease the fill volume. When the compression forceincreases, another signal tells the tablet press to fine-tune the weight cam to reduce the volume of fill. That’sall weight control is, nothing more. In short, you mustmake incremental adjustments to the weight cam toattain accurate volumetric change, which—given aknown and fixed product density—relates precisely togravimetric change.

If you’ve gotten this far and the tablet weights stillvary too much, look closely at how well the machinekeeps the product in the die before it’s compressed into atablet. Light, fluffy powders may “splash” out of the die,or the centrifugal force of the rotating turret may causethe powder to come out. Either way, the tablet will beunderweight. The best defense against splashing is to usethe “pull-down” feature offered on most tablet pressesnowadays. This feature pulls the lower cam down, allow-ing the column of powder to drop into the die after fill-ing , thus preventing under-filling. It can also prevent thepowder from splashing.

Reducing machine speed is another good strategy.Naturally, you’d prefer to move the powder through thetablet press as fast as possible, but the formulation dic-tates the speed of the press. I know companies that pur-chased a new press hoping to make their tablets faster,only to find that the formulation couldn’t match the per-formance of the press. If you want good tablets insteadof just fast tablets, you can only run the press as fast asthe product will allow. I’ve never seen a formulation thatcould outrun the machine.

Variations in tablet hardnessHardness variation is a defect, and hardness varies

according to how well you control tablet weight, pow-der density, dwell time, tablet thickness, and the workinglength of the tooling.

Increasing the tablet weight can increase the tablethardness, but your supervisor may forbid that because it“just gives product away.” Instead, you’ll be encouraged

f-Touseyart_34-43 12/1/04 11:17 AM Page 37

tabletting run timesand cleaning fre-quency. Table 1 re-peats these points.

If I were allowedjust one statementon particle size dis-tribution, it wouldbe this: Most formu-lations contain toomany fines, and by“most,” I mean about 95 percent of them. If the peopleresponsible for formulation understood compression,they would never make a product’s particle smaller thanwhat’s necessary to achieve the target dissolution rate. Ifmore products had better particle size distribution (fewerfines), the press would run better and longer with lesscleaning and fewer defects. T&C

Michael D. Tousey isowner and technicalservices director of DIPharma Tech, 152Wilkinson Drive,Westminster, SC29693. Tel. 864 6575400, fax 864 6471155. Website:

www.dipharmatech.com. Tousey has been involved in thepharmaceutical industry since 1973, including work atThomas Engineering, Shaklee, Pennwalt Stokes-Merrill,and Lakso. He founded his company in 1989. Tousey is amember of the Editorial Advisory Board of Tablets &Capsules.

I wonder if any formulators outthere have ever run a tablet press,really run a press. And I’m not talk-ing about a wimpy run that lasts 5minutes, either. In fact, I know thatno formulators have ever run atablet press under production con-ditions. If they had, I doubtwhether they’d be formulators.

Having run a press for years, Ithink the only folks more mis-guided than formulators are peoplewho think they know how to gran-ulate. Maybe we could put thesewould-be formulators in a roomwith the mill operators, throw in afew supervisors, and lock the door.Then I’d gather the people whorun the tablet press and make somereal products. There won’t be anyof this superfine stuff that doesn’tdisintegrate for 3 weeks, the stuff

that you must press so hard that itcould be used as ammo in yourshotgun. Hey, now there’s an idea.Maybe these formulators could gointo making artillery rounds, andmaybe they could hire some oftheir buddies from the granulationdepartment to go with them.

By the way, while I’m at it, I’venever, ever met a supervisor whoknows anything more than how todeliver a stack of paperwork.When it comes time to fix a prob-lem, they become politicians:“Well what do you think the prob-lem is?” they ask. It’s never said, butthe press operators must be think-ing: “You have got to be kidding!Why does the handbook say Imust consult you, the boss, whenyou don’t have a clue about whatto do?”

But back to my to my real beef:There are too many fines every-where. Why do companies spend somuch time working powders andgranulating them when they end upon the press as a pile of dust?

I’d like you formulator and granu-lator guys to get out there and lookat the issues. The fines are every-where, and it’s really a problem.Think about it: You guys cannot bemaking a good product when 30 to50 percent of it is dust. I know, theformulators will say that the fineshelp disintegration. And I’ll reply,“Baloney!” The fines are killing thepress run, they don’t compress well,and there is no benefit to fines. Youcall yourself a formulator, but any-one—you don’t need an advanceddegree—can grind powders into apile of dust. M.T.

Table 1

Excessive fines can

My gripe with formulators and granulation staff: fines!

• Remove the oil from punches, leading to tight punches that leave black specks in the tablets• Reduce the effectiveness of the formulation’s powdered lubricant, increasing the required

ejection force• Decrease yields while increasing the amount of dust collected and the frequency of cleaning• Cause capping and lamination• Decrease content uniformity and impede product flow

f-Touseyart_34-43 12/1/04 11:17 AM Page 38