This was a hot topic in Usenet news groups  in alt.drugs.chemistry way back when.  I lived near Canada at the time.  Lots of 222's up there.  8 mg of codeine plus 325 mg of aspirin.  Here's the a.d.c news group posting:

Subject:      Homebake: Codeine to Morphine
From:         dmurphy3@aol.com (DMurphy3)
Date:         1995/10/26
Newsgroups:   alt.drugs.chemistry

This came from Forensic Science International, K. Bedford, S. Nolan, R.
Onrust, and J. Siegers, v34, 1987, pp 197-204. Title is "The Illicit
Preparation of Morphine and Heroin from Pharmaceutical Products Containing
Codeine: 'Homebake' Laboratories in New Zealand".

"The O-demethylation of codeine to yield morphine has been accomplished
using a very simple method based on pyridine hydrochloride. This reagent
was introduced to opiate chemistry by Rapoport and Bonner (J. Amer. Chem.
Soc., v73, 1951, p5485) and was first applied to the conversion of codeine
to morphine by Rapoport, et al. (J. Am. Chem. Soc., v73, 1951, p5900). The
reaction was carried out under nitrogen at 220 C because morphine is prone
to decomposition or oxidation reactions at elevated temperatures and the
nitrogen atmosphere prevents or reduces these. It also limits the access
of moisture which reduces the yield."

Following is the method used, with some elaboration:
1. Crush sufficient pills to yield 2 g of codeine and mix with distilled
water. Filter with a vacuum funnel to remove insolubles and add to a
separatory funnel. Add NaOH solution to make the solution pH 12. Extract
twice with chlorofrom (2x25 mL). This will be the bottom layer. Discard
the water layer, which contains the aspirin or acetominophen) and
evaporate the chloroform layer to dryness under gentle heat. The result is
codeine base, a white crystalline powder.

2. Combine 20 mL pyridine and 25 mL conc. HCl in a beaker and heat
strongly to 190 C to drive off any water. Cover and cool rapidly to obtain
a white waxy material. This should be stored in a sealed container in the
freezer if not to be used immediately.

3. The reaction is carried out in a glass boiling tube (here one could use
a large ignition type test tube) which is sealed on one end. This should
be oven dried before use. Then 3.5 g of the pyridine salt is added to the
tube and this is then heated until it melts and for a few minutes more to
drive off any moisture. Add 1.5 g of the base and seal the tube with a
rubber stopper covered with a filter paper. Heat until the mixture begins
to fume and continue until the mixture develops a reddish-orange color and
becomes noticeably more viscous, typically 6-12 minutes.

Pour this into a 500 mL sep funnel and make the volume up to 100 mL with
distilled water. Add 10% NaOH until strongly basic. The contents will
become milky brown and then  clear brown as the solution is made basic.
When this point is reached, extract with 20 mL chloroform. This will
contain any unreacted codeine (up to 70%) and may be saved for recovery if
desired. The morphine is in the water layer.

Put the water layer in a beaker and carefully adjust the pH with HCl to pH
9 using a narrow range pHydronium paper. This is critical. Rapidly filter
using two layers of paper (here one could use a paper designed for very
fine crystals) and a vacuum flask/funnel as in step 1. A very fine brown
powder will collect on the paper. This is unwanted byproducts and should
be discarded.

Pour the filtrate into a clean beaker and, while carefully adjusting the
pH to 8.5, vigorously rubbing the inside of the beaker with a "seeding
stick" (here the authors mention that a split wooden peg is sometimes used
in the home labs in NZ; a glass stirring rod would be preferable).
Crystals should begin to form. These are allowed to settle for at least 5
minutes and then are recovered by vacuum filtering to recover the morphine
as a beige to dark brown product.

The authors report some perp's claims of 50% conversion from the codeine,
but say they obtained 30% typically, and further state that this is about
what one would expect from Rapoport's paper. Purity of up to 92% with a
more typical purity in the 80% range was reported by the authors using
this method.

TI   The illicit preparation of morphine and heroin from pharmaceutical
     products containing codeine:  'homebake' laboratories in New Zealand
AU   Bedford, Keith R.; Nolan, Susan L.; Onrust, Rene; Siegers, Jan D.
CS   Chem. Div., Dep. Sci. Ind. Res., Auckland, N. Z.
SO   Forensic Sci. Int. (1987), 34(3), 197-204
     CODEN: FSINDR; ISSN: 0379-0738
DT   Journal
LA   English
CC   4-2 (Toxicology)
     Section cross-reference(s): 1
AB   Since 1983, a large no. of small-scale illicit labs. producing
     morphine and heroin from com. available, codeine-based
     pharmaceutical products have been encountered in New Zealand.  The
     ***codeine***     ***demethylation***   procedure is based on the
     use of pyridine-HCl.  Very simple lab. equipment and reagents are
     required and these can be utilized by people with little or no chem.
     background, following a recipe-like procedure.  The process yields a
     characteristic product known as homebake.  This process is fully
     described.
ST   forensic morphine heroin prepn
IT   Legal chemistry and medicine
        (heroin and morphine prepn. in illicit labs. in)
IT   57-27-2P, Morphine, biological studies   561-27-3P, Heroin
        (illicit lab. prepn. of)

Subject:      Any synthesis, Any time, Any where - Anacharsis
From:         <102232.2645@compuserve.com (Anacharsis)>
Date:         1995/10/27
Message-Id:   <46rkg4$fk6@dub-news-svc-4.compuserve.com>
Newsgroups:   alt.drugs.chemistry

And also for all the people out
there who are constantly asking how to convert codeine to
morphine, there are numerous articles that specify *exactly*
how to do this, e.g., (the simplest one I’ve seen) A rapid, high-yield
conversion of codeine to morphine. Rice, Kenner C. Journal of
Medicinal Chemistry (1977), 20(1), 164-5. (And check out page
166, too. A comparable item.)

This improved conversion uses dimethylformamide, potassium tert-butoxide and n-propanethiol for an 80% conversion. 
The arithmetic starts looking nice with an 80% yield. I think J Med Chem is a peer reviewed journal. 
Not like a patent where no one ever tested the reaction.
J. A. Lawson, J. I. DeGraw
J. Med. Chem., 1977, 20 (1), pp 165–166

Well in that vein, why not just go codeine =Pd/C+H2=> dihydrocodeine =HBr/AcOH=> Dihydromorphine ?

Oppenauer oxidation of that will give dihydromorphinone and you'll be happy as Larry in no time...

Trouble is - not a lot of people have the means to carry out catalytic hydrogenation (although this one seems nice) and HBr is a "small" problem here (can be solved).

Oh yeah, before anyone thinks this is pie in the sky time - here's the paper where the hydrogenation/demethylation is detailed (with yields)

Practical and High-Yield Syntheses of Dihydromorphine from Tetrahydrothebaine and Efficient Syntheses of (8S)-8-Bromomorphide

Anna K. Przybyl, Judith L. Flippen-Anderson, Arthur E. Jacobson, and Kenner C. Rice (See that last name?)

J. Org. Chem., 2003, 68 (5), pp 2010–2013
DOI: 10.1021/jo0206871

I have been reviewing this procedure, and found the following references:
"Dihydrocodeinone (hydrocodone) can be demethylated with 48%aq HBr to yield hydromorphone
J. Org. Chem. 3, 204 (1938)
9g Ethyldihydrocodeinone in 45ml 48% HBr, reflux 14 min, dilute with 100ml water and basify with NaOH. Wash with ether to remove non-phenolic material, add excess ammonium chloride and extract with 4000ml diethyl ether [in at least two dozen portions], evaporate the ether and recrystallize from alcohol to give 8.4g ethyldihydromorphone 48%aq HBr to demethylate the methoxy ether on about a half-dozen alkyl dihydrocodeinone derivatives via a 20-30 min. reflux.
"All yields are over 95%"

“Dihydrocodeine from Codeine:
In a Parr hydrogenator jar, codeine (2.0g, 6.7 mmol) and 10% Pd/C (200mg) were combined in ethyl acetate (75ml). This mixture was hydrogenated at room temp at 37 psi for 2h. During this time, the hydrogen pressure decreased to 34 psi within 30 min, was increased back to 37 psi, and thereafter remained constant for the duration of the reaction. The mixture was filtered through Celite, the filter cake washed with 2x10ml ethyl acetate, the resulting solution stripped of solvent using a rotary evaporator by high vacuum. The product foamed considerably upon initial application of high vacuum, so care was taken not to lose any product. Leaving the product on high vacuum overnight afforded a white, crunchy, crystalline solid (2.0g (99.3%), mp 113-4°C [lit. 112-3°C]).”

As well as:
"synthesis of essential drugs"

excerpt:
"Hydromorphone: Hydromorphone, 4,5-epoxy-3-hydroxy-N-methyl-6-oxomorphinane
(3.1.22), is a compound related to morphine that differs in the absence of a double bond between C7–C8 and the presence of a keto group instead of a hydroxyl group on C6. The drug is synthesized by the isomerization of morphine in the presence of a palladium or platinum catalyst [11,12]. The other way is by oxidation of dihydroporphine [13,14]. Hydromorphone is more soluble than morphine and approximately eight times more active
upon parenteral administration. High solubility permits a lower volume of injected fluid, which is important if multiple injections are needed. It begins to work faster than morphine, but lasts for a shorter amount of time. It has a high sedative effect and a lessened capability of causing euphoria. Hydromorphone is used the same way as morphine. Side effects are analogous. Synonyms for this drug are dilaudid and others."

The same procedure works for hydrocodone substituting codeine for morphine. According to the opiate/opiod equavelence chart the hydro-derivatives of these are 5 times more potent then there parent compounds.


A palladium catalyst isn't terribly hard to come by.  PdCl2 isn't suspicious in small quantities and can easily be reduced to palladium black which would be suitable.  As for the HBr, could you not do a simple replacement with NaBr in HCl.  NaBr is sold in bulk cheap for "bromine pools".  Chlorine should have a greater affinity for the sodium ion.  Then some distillation should get the HBr or so I would think.

Another comment for catalysts; catalytic converters are ripe with palladium, platinum and ruthenium complex's.  I'm curious if there is any literature detailing which complexes, but I believe catalytic converters from scrap could be a good place to look for cheap catalysts.

This procedure is interesting because of its seeming ease, short reaction time, and high yield.  Sure there are routes to more potent material, but  mg doses give a certain wiggle room for administration rather then risking uncertain doom with very potent material.

Thoughts?

I found two references of interest.

xxx.resources.metapress.com/pdf-preview.axd?code=nw14532h43182226&size=largest

In this source I am very curious if DMSO could be substituted for dimethyl sulfide?

the other:

'A practical High Yielding Synthesis of Dihydromorphine from Tetrahydrothebaine & Efficient Syntheses of of 8(S) 8 -Bromomorphide' by Anna K. Anna K. Przybyl, Judith L. Flippen-Anderson, Arthur E. Jacobson and Kenner C. Rice of NIH.

My question would be can you go from dihydrocodone to dihydromorphone via the Hbr/acetic route or does it only work with dihydrocodeine to dihydromorphine.  I don't see any reason why it shouldn't.