Manipulation and Flow of Biological Fluids in Straight Channels Micro Machined

8/6/2019 Manipulation and Flow of Biological Fluids in Straight Channels Micro Machined

http://slidepdf.com/reader/full/manipulation-and-flow-of-biological-fluids-in-straight-channels-micro-machined 1/5

CUN . CHEM.40/1,43-47(1994) #{149}utomation an d An al yt ic al T ec hn iq ue s

CUNICAL CHEMISTRY,Vol .40, No. 1, 1994 43

Man ipu la tion and F low o f B io logica lF lu ids in S tr aigh t Channe ls M ic romachined

in Si licon

Peter W ild ing,’ Joseph Pfahier,2 Haim H . Bau,3 Jay N . Zem el,4 and Larry J. Krlcka”5

Analysis of m in ute s amp le v olume s is a m ajo r analyt ical

chal lenge th a t r equ ir es an unde rs ta nd ing o f flu id f low inmic ros tructu res . Accord ing ly , f low dynamics o f b io log ica l

f lu ids and cell suspensions in s tr aigh t glass-capped sili-

co n microchannels (40 to 150 m wide, 20 and 40 m

deep) we re s tu die d. W e demonst ra ted tha t these micro -structures a re a pp ro pr ia te components fo r microfluidic

analy tic a l devices. D iffe re nt flu id s w ere easily manipu-la te d in th e m ic ro ch an ne ls , a nd me as ur ements o f flow

ra te as a function o f p re ss ure fo r w ho le h um an blood,

se rum, p lasma , and ce ll suspensions reveal ed non -New-

to nia n b eh av io r. B y me an s o f m ic romac hin ed f ilte rs (5

m) located i n channe ls , b lo od c ells a nd m ic ro pa rtic le s

were e ff ec ti ve ly sepa ra ted f rom nano lit er -s ized samp les ,

c lea rly i nd ica ti ng the future ro le o f m icros truc tu res for av ar ie ty o f a na ly tic al p urp os es .

Indexing Terms: viscosity/shear rate/non-Newtonianfluids/micro-

channels

A trend in clinical laboratory analysis has been th e

progressive reduction in the volume of sample required

fo r assay. Early methods required at least 1 mL of sam -

p le , b ut modern-day tech niques can analy ze volumes as

small as 1 L (1). Various factors a re r es po ns ib le fo r

th is trend, including concerns over th e hazard po sed by

bloo db orne patho gen s in some clinical sam ples (hum an

immunodeficiency virus; hepatitis A , B , and C ), pa-

t ie nt s’ c on v en ie nc e (fingerstick samples), reduction in

reagent usage, and cost of testing. The n ex t a na ly tic al

challenge is to further reduce sample volumes to the

nanoliter-p icoliter range (10_9_10_12 L). M any clini-

cally im portant substances are present at micromole per

liter con cen tration s (10-s m ol/L ); thu s, 1 nL of sample

could contain 1 fmol of analyte, w hich corresponds to

100 m iffion molecules. This num ber of molecules is well

within th e detection range of several ana lytical m eth od s

(e.g., chemiluminescence an d fluorescence) (2). The m a-

jo r hurdle to implementing analysis on nanoliter an d

picoliter volum es is that th e m ic ro flu id ic s r eq uire d fo r

handling such small volumes has not been developed.

We b elie ve th at m icro ma ch in ed sfficon structures in -

corporating in tercon nected m icro ch ann els an d cham-

b ers w ill p rov ide th e necessary devices for hand ling and

analyzing very small volumes of sample (3 , 4 ). Silicon

m icrom achining has been used to produce m icrome-

Departments of ‘Pathology a n d La bo ra to ry Medicine, ‘Me-

chanical E ngineering and A pplied M echanics, and4 E lectrical En-g in e er in g , Un iv er si ty of P enn sy lv an ia , P h il ad e lp h ia , PA 19104.‘APD C ry og en ic s, A lle nt ow n , PA 1 81 03 .5 Au th or f or c or re sp on de nc e. F ax 215-662-7529.ReceivedJune 16 , 1993 ; a cc e pt ed S ept ember 10 , 1993.

chanica l devices such as valves, filters, cantilevers, ro-

tors, and gas chromatographs (5-9). This technology is

read ily ada pta ble to the development of microfluidic

analytical devices. W e u sed p hotolith ograph ic tech-

n iques to fabricate straight channels, 20-40 m deep ,

an d filter system s of different geo metries on silico n

ch ip s. The advantage of this new test system is that

disposable, highly reproducible, complex microchannel

networks can be batch-manufactured inexpensively.

We report here the flow characteristics o f b io lo g ic a l

flu id s a nd blood cell suspensions in straight channels of

differing size with volumes in the nanoliter range.

Mate rIa ls a nd Methods

Fabrication of microchannels. Microchannels and m l-crochannel structures (Fig . IA ) we re f ab ric ate d by using

planar p ho to lith og ra ph y o r reactive io n etching in sili-

con (400-sm -thick wafers, cry stallog ra phic orien tation

<1,1,0>, anisotropic KOH etchant) (SGS Thomson,

M ontgom eryville, PA) w ith photolithographic m asks

(Align-Rite, Santa Clara, CA ). The wafers w ere diced

into 17 x 11 mm chips and then sealed w ith 1 .58-mm -

thick P yrex glass (Mooney P re cis io n G la ss , H u ntin gto n,

W V) by using a diffusive bonding technique (10). An

entrance and an exit port (500 x 500 m ) were etched

through the 400-pin-th ick silicon at each end of the

11.7-m m-long channel. For each batch of channels, the

depth of one channel was measured w ith a surface pro-ifiom eter w ith a precision of ± 1%. C han nels p rod uced

w ith th e an isotrop ic etchant had a trapezoidal cross-

section (F ig. 1B ).

Flow studies. We quantitated f lu id flow in m ic ro ch an -

nels w ith th e experim ental apparatus shown in Fig. 2. A

chip was held in a custom-built holder (F aulkner Instru-

ments, Pitman, NJ) on a m icroscop e stage (Aristomet,

Wild L eitz, H eerbrugg, S witzerland). The in let port of

the ch ip was connected to a syrin ge, housed in a syrin ge

driver (M odel 350; Sage Instruments, B oston, M A ), an d

actuated by a stepper m otor (m arimum force 20 N ). A

load cell m easured the applied force from which th e

d riv in g p re ssu re w as d edu ced . The f or ce m e as ur em e nt

was corrected for frictional lo sse s in th e system, i.e ., d ue

to f ric tio n b etw ee n th e s yr in ge plunger and the syringe

barrel an d pressure losses in th e s yr in ge itself. The force

req uired to overcom e these fr ic tio na l lo ss es w as d eter-

mined by filling a syringe w ith distilled water and then

measuring th e force required to empty the syringe as a

function of flow rate. The fun ction al losses rang ed from

5% to 20% of the tota l force at the h ighest to the lowest

pump speed, respectively. The flow rate for each of the

speeds was obtained by t iming th e syringe plunger dis-

placem ent. A ll m easurem ents were conducted under a



Flow Channel- \B

Inlet Sump

SIllcon-i

AA

LOADca

1=c

I YRDRNER

Fig . 2. Expe rImen ta l a pp a ra tu s fo r o bs er vin g a nd recording flo w Insi licon microchannels .

VDU , v i d eo d is p la y u n it ; VCR, videocassette recorder.

Out let Sump

A

4 4 C UN IC ALCH EM IS TRY,Vol. 4 0 , No . 1 , 1994

a. Top View

Row Channe l

Inlet Sump - Silicon Out let Sump

b . C ro ss -Sec tio na ’V iew (Sec t. A -A )

c . C hann el C ro ss -S ec tio n (S ec t. B -B )F ig . 1 . (A ) Silicon chip w ith tw o straigh t channels (airows) (upperchannel 60 ian wide, low er 80 cufl wide); (B ) Schematic of s i li conchip.

m icroscope to detect channel clogging. The pressure

losses in the syringe and supp ly p ip ing was <1% of th e

to tal pressure drop.

Flow in the channels was observed and recorded w ith

a black an d w hite television cam era (D age-M TI, M ich-

igan City , IN ) and a videocassette recorder (P VM -122;

Sony, Teaneck, N J). No failures were detected in th e

diffusive bonded glass-silicon junctions in an y of the

chips studied. A ll studies w ere perform ed at am bien t

temperature (22.3 ± 1.5#{176}C)n a n e nv iro nm e nta lly con-

t ro ll ed l ab o ra to ry .

Biological samples and m icrob ead s. Blood from a

h ealthy v olun teer (J.P .) w as collected into heparinized

and plain tubes. A leukocyte (W BC) suspension was

prepared b y u sin g Lymphocyte Separation Medium (O r-

ganon Tekn ika, Durham , NC) (11). Washed erytbro-

cytes (RBCs) were prepared with isotonic sa lin e. L atex

m icrobeads (5 .78 pin diameter) were purchased from

Polysciences (Warrington, PA).

P rotein -co ated cha nne ls. The inside surfaces of some

of the microchanneis were coated w ith albumin b y fill-

ing them with a solution of albumin (5 0 g/L ) (Sigma, St .Louis, MO), and then drying the channels in a m icro-

wave oven .

Results

We investigated the flow of complex fluids (whole

bloo d, serum , washed RBCs an d WBCs) from a 50 0 x

500 m entrance sump into straig ht c han nels (11.7 mm

long) w ith cross-sectional dimensions 80 &m wide (top)

x 20 m deep (volume -17 nL) an d 150 m wide (top)

x 40 p in deep (volume -64 nL ). The results of ou r

measurements are depicted in Figs. 3 and 4 for the

40-pin- and 20-pin-deep channels, respectively . F igs. 3A

and 4A show thepressure

head (thepressure

needed to

drive the flow ) as a function o f f low rate. Figs. 3B and 4B

dep ict the pressure head needed to transport various

biological fluids (APB) norm alized w ith th e pressure re -

quired to transport w ater CAP) at the same flow rate

(relative viscosity &R = P3/iP).

Although it is conven tional in flu id mechanics topre-

sent resu lts as functions of th e Reynolds n um ber (c ross-

sectional averaged velocity x effective diameter/viscos-

ity), w e have not done so because of possib le ambiguity

in the definition of th e viscosity of biological fluids that

e xh ib it n on -N ewto ni an behavior.

150-pm-wide x 40 -pin -deep ch an nel. Flow rate as a

function of pressure head and relative viscosity for d is-

tilled water, serum , and RBCs are shown in Fig. 3. The

p ressu re re quire d to ach ieve a given flow rate increased

as the complexity an d viscosity of th e flu id increased -

e.g., for a flow rate of 0.25 mLlm in , distilled water re-

quired a pressure of-75 kPa, whereas th e identical flow

rate w ith whole blo od requ ired >250 kPa (Fig. 3A) .

Whole blood,, diluted 3500:1 in iso to nic sa lin e to give th e

same cell concentration as the WBC preparation, re-

quired less driving force than th e WBCs at a given flow

rate. T his is because whole blood is made up primarily of

RBCs, which are highly flexib le an d much smaller than



A

I50 0

45 0

40 0

35 00

a-- 30 04)

25 0(4

20 0

a- 15 0

10 0

50

I

1 60 0

1 4 0 0

I

00.03

o D istilled Woter - -#{149}Serum - - -

A RBC in s al in e - - -

6. 5

0.09 0.15 0.22 0.28 0.34 0.40 0.46 0.53 0.59 0.65F lo w R ot e, m L/m in

RBCSol ine

CLINICA LCHEMISTR Y,Vol. 40 , No . 1 , 1994 45

5.5

2’45

UU,

53.5

>0

, 2. 5

1 .5

0. 5 -

0.05

p1200

- 1000

(44)

a-

0.15 0.25 0.35 0.45 0.55 0.65

F low Ro te. nL /m in

Ag. 3. (A ) P ressure head and (B ) r ela tiv e v is co sity a s a f un cti on o fth e flow r at e f or different fluids in a 150-m -w ide x 40-1an-deep

channel.

W BC s (m ean corpuscular volume, 87 fL vs 230-470 fL)

(12).

For distified w ater, the Reynolds num ber ranged from17 to 126, and the experim ental results were in good

agreement with theoretical predictions (solid line, F ig.

3A ) obtained by solving th e Navier-Stokes equations

(4). The theory p re dic ts lin ea r dependence between th e

pressure head (SF ) and flow rate (F ) according to the

equation -“w = AF, in wh ich w is the viscosity of

water and A is a constant that depends on ly on channel

geometry. The increase in pressure drop above theoret-

i ca l p re d ic ti on s at the h igh flow rates is due to inertial

losses an d en tra nc e len gth effects, which were not in -

duded in the theory. The shear rate, wh ich is propor-

tional to the mrnrimum velocity in the cross-section di-

vided by the half-height of the channel, ranged from

3086 to 39439 s’. Because of the l imitations of the

syringe pum p, the m ore viscous fluids w ere no t tested at

the h ighest flow rate; and because of the inaccuracy of

th e load cell at low forces, th e distilled water was not

tested at the lowest flow rate.

Saline (15.4 mmol/L) behaved alm ost identically to

distified water in the channel. This imp lies that the flow

of liquids in these sm all channels is not affected by the

presence of electro lytes. In spite of th e large surface-to-

cro ss-sec tio na l a re a ratio, electrostatic forces d id not

appear to play an important role in our experiments.

B

80 0

60 0

40 0

20 0

00.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.10 0 .11 0 .12 0.13

F lo w R ot e, m L/m in

3.15

3.00

>2.85

2.70

2.55

. 2.400

4)

a - 2 .2 5

2.10

1.95

1.800.045 0.055 0.065 0.075 0.085 0.095 0.105 0.115 0.125

F lo w R ote , m L/m in

Fig. 4. (A ) P ressure head an d (B ) re la tiv e v is co sity a s a fu nc tio n o fflow rate fo r d i st il led wa t er , se rum, an d R BC suspension in saline ina 80-L m-w tde x 20-1inI-deep channel.

For the shear rates in our experiments >3000 s’, all

th e b iolog ical fluid s ex hib ited the approxim ately linear

relationship 2B = -‘O ,B + i5A1 in wh ich PB is the

pressu re head needed to drive th e biological flu id and 1B

can be interpreted as th e viscosity o f t he b io lo gic al fluid.When the flow rate F is extrapolated to zero, then the

pressure head (AP0) is not zero as in the case of New -

tonian fluids (i.e ., water). Th is behavior is sim ilar to

that of a Bingham plastic (a flu id that re qu ires fin ite

yield stress before it begins t o f lo w) .

T he r ela tiv e viscosity [P R = P/(AF) + ‘w]

i s d ep ic te d in Fig . 3B as a function of flow rate (F). For

Newtonianfluidssuchassaline,APOB = 0andRisa

constant. Thus, as anticipated, th e r ela tiv e viscosity of

saline was constant over the range of flow rates tested

(Fig. 3B). For b io log ica l flu id s at high f lo w r ate s, 1R -

FLat/Lw= a constant, whereas at very low flow rates, FL R

- P/(,AF); i.e., ILR i s i nve rs e ly proportional to F ,

an d as F - 0, ILR - infinity. Thus fo r serum, plasma,

and cell suspensions, th e rela tive viscosity increased

markedly as the flow rate decreased (<0.35 mI.dmin);

e.g., viscosity of w hole blood increased more than two-

fo ld as flow rate was reduced from 0.35 to 0.1 mL/min

(Fig. 3B).

80-pm-wide x 20-pm-deep channel. The relation be-

tween flow rate and relative viscosity and pressure for

distified w ate r, seru m, an d RBCs in saline is dep icted in

Fig . 4A and B. For distilled water, the Reynold s num ber

varied from 20 to 50 an d th e shear rate in the channel



Ag. 5. F il tr at io n o f 5 .7 8 -am -d iamet er b e ad s ma 5 00 -tm -wi de c ha n-nel by a f il te r w it h 5 -t ni spacings.

46 C UN ICA L CH EM ISTR Y, Vol. 40, No. 1 , 1994

ranged from 20 800 to 105 257 s’. The b iological flu id s

required a greater pressure than d id distilled w ater to

achieve th e same flo w r ate (F ig . 4A). Compared with

that for the larger channel, th e p re ssu re re qu ired to

achieve a given flow rate w as m uch higher; e .g ., a se rum

flow rate of 0.09 mL/m in required a pressure of 1100

kP a in th e sm all channel and 150 kP a in the large

channel. B oth se ru m an d RBCs in s al in e e xh ib it ed non-

Newtonian behavior (i.e ., the relative viscosity in -creased as the flow rate decreased) (Fig. 4B). Flow of

whole blood in this channel was not stud ied b ecau se of

clogging o f th e channel.

Albumin-coated 150-pm-w ide x 40-pm-deep chan-

nels. We a ls o s tu die d the flow of WBC s suspended in

isotonic saline (1 .8 x 106 cells/L) in a 150-pm -w ide x

40-pm-deep channel and th rough an identical channel

coated with albumin. The coating was uneven and we

were not able to measure it s thickness. However, after

viewing the coating under a m icroscope, w e estimated

that the co atin g’s p eak s were no more than 5% of the

channel d ep th . W e detected no difference in th e pressure

drop vs flow rate measurements of the WBCs flo wingthrough th e coated an d uncoated channels, nor did we

observe any tendency for the WBCs to attach to th e

album in (they tended to stay away from th e walls).

Microfilters. Filters of differing designs were fabri-

cated by reactive io n e tc hin g w ith in channels to study

separation and (or) iso lation of form ed elem ents (e.g.,

b lo od c ells ). The ability of in div idu al ifiter designs to

remove p ar tic ula te m ate ria l from test solu tion s w as in -

vestigated with suspensions of beads and RBCs . The

5-pin filter show n in Fig. 5 comprises a row of 12 com -

plex posts that form a barrier across th e channel. This

filter effectively separated latex microbeads (5 .78 pm

diameter). It also separated RBCs, but the separationwas less complete because of the deformability of the

cells that allowed som e cells to pass through th e filter.

We did not observe any destruction of RBCs in th e

channels , but in other studies (not reported h ere ) c on -

ducted at higher pressures a nd flo w rates than th e on es

used here, w e d id o bs er ve significant RBC deformation.

DIscussIon

W hole blood is a shear-thinning (the apparent viscos-

ity decreases as th e shear rate increases) non-Newto-

nian fluid . B lood viscosity depends on cell content,

plasma p rot ei n con c en t ra ti on , temperature, storage con-

ditions, health of donor, tim e since eating or exercising,

and testing conditions. For healthy males, th e nominal

whole blood viscosity is in th e r an ge 3 .6 2-5 .5 6 mPa at

37#{176}Cnd a shear rate of 230 s (12). At shear rates>1000 s’ andintubeswithdiameters > 22 pm ,b lo od is

believed to behave like a New tonian l iquid; but at low er

shear rates, it behaves like a Bingham plastic (a fluid

that requires a fin ite yield stress before it begins to

flow-e.g., toothpaste) (13,14). At low shear rates, blood

exhibits a sm all but fin ite yield stress that c an a ffe ct th e

velocity p rofile a nd p ressure d rop in sm all tu bes at low

flow rates. The viscosity o f whol e blood and suspens io n s

of RBCs exhibits a dependence on the size of the tube

when the d iameter is <300 pm (Fahraeus-Lindqvist

effect) (15)-e.g., t he r el at iv e viscosity of normal hiimsin

b lo od d ro ps by a factor of two when the diameter of th e

tube is reduced from 500 to 40 pm (15). W e observed theFahraeus-Lindqvist effect for R BC suspensions in th e

tw o c ha nn el sizes stu died (F igs. 3B and 4B ). T he relative

viscosity o f t he RBC s us pe ns io n, at a constant flow rate of

-0.05 mL/min , changed from 4.5 to 2.6 as th e channel

s iz e w as reduced from 40 x 150 pm to 20 x 80 pm .

Our results suggest that sim ple liquids behave both

qualitatively and quantitatively according to estab-

lished theories (from their m acroscopic counterparts) in

channels w ith characteristic d imensions >20 pm . In

spite of the large surface-to-volume ratio (-10 m21m3) ,

we w ere unable to detect an y s ig nif ic an t in flu en ce on

th e test results by s urfa ce f or ce s. In contrast to sim ple

flu id s, b iological flu id s exhibited shear-th inn ing, non -N ew to nian beh avior. T he shape of the curves in Figs. 3B

and 4B are as expected for a p la st ic s he a r- th in n in g fluid.

H owever, w e observed this behavior at shear rates be-

tween 3000 and 20000 s’, whereas in o th er s tu die s

this was observed only a t flow rates <1000 s’ (1 3, 1 4) .

T he r ela tiv e ease of handling of the flu ids in the

microchannels indicates that this type of system will be

appropriate for the flu id ic componen t of an analytical

d ev ic e d es ig ne d to test nanoliter-picoliter sample vol-

um es. W e are currently exploiting m icrofabrication

technology for the construction of inexpensive, d ispos-

able diagnostic devices fo r medical app lication s. A fur-

ther app lication of the m ic ro fa bric ate d stru ctu re s is

with experimental analogs fo r th e study of transport in

c omp le x n e two rk s and interconnected pores. M icrochan-

nels, such as capillary networks, play an important role

i n b io lo g ic a l processes, but detailed study of flu id flow

has been ham pered by the lack of convenient or versa-

tile experim ental system s. P revious flow studies have

used either direct in v ivo observation of microvessels

(e .g ., r ab bit e ar , hamster cheek pouch) (1 6, 1 7), flow in

thin glass tubes (18), or microchannels made by casting

th in w ires in plastic o r silic on e ru bb er (19). T he m ic ro -

fa bric ate d s tru ctu re s offer a versatile alternative to



CUNICALCHEMISTRY ,V ol. 4 0, N o. 1 , 1994 47

these test system s, an d recently a similar system has

been used to s tu dy b lo od rh eo lo gy in arrays (num ber =

2600) of very short (14.4 pm ) triangu lar cross-section

channels (equivalent diam eter, 6 pm ) (20).

References

1 Cassaday M , D iebler H , Herron R , Pelavin M , Svenjak D ,

V lastelica D . C apsu le ch em istry tech no lo gy for h igh -speed chem-ic al a na ly sis . C li n C h em 1 98 5;3 1:1 45 3- 6.2. K ricka L I. Chem ilum ineacent and biolum inescent techniques

[Rev iew ]. C li n Chess 1991;37:1472-81.3. K ric ka L 1, W ild in g P , P fa hl er J, Harley J, B au H , Zemel JN .Liquid transport in m icron and sub micro n c han nels. S oc P ho to -O ptic al I ns tr uin E ng 1989;1167:159-68.4. Pfabler JN . Liquid transport in m icron and subm icron size

chan nels [D issertatio n]. P hiladelph ia, P A: U niversity of Pennsyl-vania, 1992:lO9pp.5. AngeII JB , Terry SC, Barth PW . Silicon m icromechanical

d ev ices. S d A m 19 83;2 48 :4 4-5 5.

6 . R apoport SD , Reed M L, W eiss LE. Fabrication a nd testin g o f am icrod ynam ic roto r fo r blood flow measurements. J Micromech

Microeng 1991;l:60-5.7. K ittilsland G , S tem me G, Norden B. A sub-micron partidefilte r in s il ic on . S ensor s Ac tu at or s 1 9 90A21-A23 :9 04-7.8. W ise K D, N ajafi K . Mi cr of ab ri ca ti on t ec hn iq ue s fo r integratedsensors an d m i crosys tem s. Sc ie nce 1991;254:1335 -. 42.9. Mallon J. N an ote cb no lo gy fr om a mi cr oma ch in is t’ s p er sp ec ti ve .

In : C ra nd all BC , Lewis J, ads. Nan ot ec hn ol og y. R es ea rc h an dperspectives. Cam bridge, M A M IT Press, 1992:215-40.

10 . W allis 0. D irect-current polarization during field-assisted

g la ss -m e ta l s ea li ng . JA m Ceram Soc 1970;53:563-7.11 . Fotino M , M erson EJ, A llen FH . M icromethod fo r rapidseparation of lym ph oc yte s from pe riphe ral blood . A nn C lin L ab S d1971;1:131-3.12 . Lentner C , ed Geigy scientific tables, VoL 3. Basel: Ciba-G ei gy , 1 98 4:6 8.

1 3. W hitxn ore R L. R heolo gy o f t he circ ula tio n. N ew York: Perga-mo n, 1 96 8: 66 -7 0.

14 . Skalak R , O zkaya N , Skalak TC . B ioflu id mechanics. Ann RevFluid Mech 1989;21:167-204.

15 . M idd lem an S . T ran spo rt phenomena in t he c ard io v as cu la rsy stem . N ew York: W i le y, 1 97 2: 29 9p p.

16 . Schmid-Schonbein GW , Skalak R , IJsam i 5, Chien S . Celld istribu tion in cap illary n etw ork s. M icro vasc B aa 1 98 0;19 :18 -44 .

1 7. Ma yr ov it z HN , R ub in R . L euko cy te d istrib ution to arteriolarbranches: d ep end en c e on mi crova scu la r blood f low . M i crova sc Re s1985;29:282-94.18. La Celle PL Alterations by l eu ko cy te s o f e ry th ro cy te f low inmicrochannels. Blood C e ll s 1 9 86 ;1 2 :1 7 9- 89 .

19 . Fenton BM , Carr RT , Cokelet GR. Non-uniform red celldistr ibution in 20 to 100 m icron bifurcations. M icrovaac Rea

1985;29:103-26.

20 . Kikuchi Y , Sate K , Ohki H, K an ek o T . O ptically accessib le

microchannels formed in a s in g le .c ry st al silicon substrate for

stu dies o f b loo d rheo log y. M icro vasc R es 1 99 2;44 :22 6-4 0.

Manipulation and Flow of Biological Fluids in Straight Channels Micro Machined

Documents

Transcript of Manipulation and Flow of Biological Fluids in Straight Channels Micro Machined