X-Ray Analysis of Apple Products

...not kind of x-ray. these soft x-rays, within 40 kev (whereas doctor x-rays go 110 kev or more). aren't looking through macbooks or iphones, rather 'seeing' elements. these x-rays cause elements compose istuff fluoresce color, result can see apple uses make things.

first - note @ looking at: x-ray fluorescence (xrf) spectra. try think of these color peaks. example, color red 1.7 electron volts (ev). color violet 3.1 ev. color iron 6,400 ev, or 6.4 kev. each element has color on spectrum - , looking at. height of each peak controlled number of factors (energy send in, fluorescence efficiency, measurement depth, etc.) here our comparisons make use of factors. means when see difference in peak height, looking @ change in number of atoms.

secondly, our analysis qualitative. because doing non-destructive analysis, @ mercy of anodization, thin glass layers, , other trickery in mac products.

- analysis done using rmb running parallels , hooked hdtv (i wanted test new vga-usb-c adapter came in). wondering, did fine old rmbp did.

said, first question. going on aluminum in apple products?

aluminum

green line on counter aluminum foil control (reynolds wrap). red line new gold retina macbook (rmb, 8,1). first - notice height of aluminum peak in foil in comparison macbook? both made of aluminum alloys (al), happening macbook effects of anodization, bath in sulfuric acid. can see sulfur peak (s) right high aluminum peak. interpretation - looks there mixed layer of sulfur , aluminum on surface of macbook. when touch macbook, or iphone, or ipad, aren't making contact aluminum layer of sulfur oxides.

can shift gears , take @ trace elements present :

reynolds wrap has fair bit of iron (fe), give bit more strength , color. gold rmb has lot more nickel (ni) , little more copper (cu). both have gallium (ga)- uninentional. gallium sits below aluminum in periodic table , has same number of valence electrons in outer orbital. gallium is, in short, hitchhiker along ride.

apple’s istuff comes in 3 different colors, silver, space gray, , gold. differences between these colors? used silver iphone 6, space gray ipad mini, , same gold retina macbook.

answer not much, subtle variations give clues how apple achieves colors. first, note space gray ipad has more iron (fe) in other two. reduced iron has black color, while oxidized iron has red color (rust). swag (scientific wild ass guess) reduced iron being used achieve black color. note space gray ipad , silver iphone both have same amount of nickel (ni). gold rmb quite different - has more copper (cu) , more… …manganese (mn)? that’s strange. not sure going on here. manganese oxide has purple color, , 1 of first color additives used in bronze age glass. used maya dark pigment temples. perhaps here being used mute copper give more orange, gold-like look. lost - maybe out there more familiar metallurgy and/or anodization can shed light.

speaking of light, apple has been shifting using metal logos in place of backlight. can see made of.

metallic apple logos

here, comparing gold rmb piece of stainless steel - picture. apple logo stainless steel, has more nickel (ni) , titanium (ti) relative reference material (a knife kitchen). can compare apple logos 3 color types, silver, space gray, , gold:

silver iphone 6 used bit more chrome , nickel shift shade. gold rmb has more titanium (ti) no copper (cu) - color affect achieved differently in aluminum. space gray, pattern far, using more iron (fe), in reduced state. things more interesting if @ higher energies:

have no idea zirconium (zr) doing in space gray anodization option, or why totally absent gold , silver. haven’t seen before, , analyze lots of weird stuff. molybdenum (mo) see common additive iron alloys, pretense not surprising, though higher in silver iphone in gold rmb. not sure make of - perhaps additional molybdenum helps achieve color going for.

perhaps interesting details come turning x-rays onto glass.

glass

let’s focus on iphone first. corning revived gorilla glass concept original iphone, though first phones had unwelcome amount of arsenic (yes, it’s come when analyzing old iphones). have been arsenic free while - how did attain strength? answer in potassium (k). 1 of strongest forms of glass, , 1 of oldest. first glass made in part adding volcanic ash clay melted @ high temperatures - volcanic ash had lots of potassium. corning did invent new king of glass (kinda), borrowing deep history of glass making. sediment source glass had lots of aluminum (al) , silicon (si).

things weird when @ rmb glass. first - low silicon (si) - @ first thought weren’t looking @ glass @ all, x-rays behaving if glass (too little inelastic scatter plastic). however, think iodine (i) peaks right give answer - glass coated iodine. i’m not sure - explain low silicon (si) peak since low energy , have tough time getting through glass. put a guide if you’re interested in attenuation of these high-energy photons. long story short, elements aluminum, silicon, or phosphorous analyzing 40 or microns (0.001mm) deep @ most. calcium trace has come along whatever sediment using make glass in rmb. phosphorous may part of coating. in fact, has to show glass.

if shift higher energies, things weird.

that’s lots , lots of strontium (sr) seeing. strontium can serve strengthening agent, may seeing here; more strontium used in rmb add strength not gorilla glass. if go higher energies, things weirder.

don’t work part of spectrum often, , should have used heavier duty filter. can still learn this. first, tin peak (sn) indicates method of manufacture. after speaking various people smarter i, seems pouring glass mixture onto small pond of liquid tin. glass cures @ lower temperature tin, still liquid. helps create ultra-flat surface, critical touch input , precise fitting. but, little tin comes ride. element did not expect indium (in) - perhaps indicate igzo (indium, gallium, zinc oxide) screen? not sure, because gallium (ga) , zinc (zn) present in both iphone , macbook glass. also, analyzed retina macbook pro late 2012, , has indium (in). measuring element @ quite depth since higher energy, hitting component within screen well. way know sure remove 1 of these screens , measure other side - not willing shiny new rmb.

technical details
feel free skip if not interested in how analysis conducted. first, bruker tracer iiisd used - has 2w rhodium anode silicon drift detector (sdd) capable of full width height maximum (fwhm) of 150 ev count rate of 200,000 photons per second, though these measurements fell bit below (which better, actually). used 40 kev combined 30 μa - multiply 2 , using 1.2 watts, as electric screwdriver. light elements used no filter - using xrf instrument scanning electron microscope without electrons. trace elements zirconium , molybdenum, used 25 μm cu/25 μm ti/300 μm al filter. cut out lot of background noise clearer reading on elements.

count rate filtered measurements around 60,000 photons per second, while went high 150,000 counts no filter measurements. latter should have had lower current, since qualitative analysis didn’t mind. each spectrum took 60 seconds accumulate. more time better, arm started hurt holding macbook glass.